Showing 1–50 of 73 results for author: Abreu, S

An Algebraic Approach to Weighted Answer-set Programming

Authors: Francisco Coelho, Bruno Dinis, Dietmar Seipel, Salvador Abreu

Abstract: Logic programs, more specifically, Answer-set programs, can be annotated with probabilities on facts to express uncertainty. We address the problem of propagating weight annotations on facts (eg probabilities) of an ASP to its standard models, and from there to events (defined as sets of atoms) in a dataset over the program's domain. We propose a novel approach which is algebraic in the sense that… ▽ More Logic programs, more specifically, Answer-set programs, can be annotated with probabilities on facts to express uncertainty. We address the problem of propagating weight annotations on facts (eg probabilities) of an ASP to its standard models, and from there to events (defined as sets of atoms) in a dataset over the program's domain. We propose a novel approach which is algebraic in the sense that it relies on an equivalence relation over the set of events. Uncertainty is then described as polynomial expressions over variables. We propagate the weight function in the space of models and events, rather than doing so within the syntax of the program. As evidence that our approach is sound, we show that certain facts behave as expected. Our approach allows us to investigate weight annotated programs and to determine how suitable a given one is for modeling a given dataset containing events. △ Less

Submitted 28 March, 2025; v1 submitted 26 March, 2025; originally announced March 2025.

MSC Class: 68M01 ACM Class: D.3.1

Neuromorphic Principles for Efficient Large Language Models on Intel Loihi 2

Authors: Steven Abreu, Sumit Bam Shrestha, Rui-Jie Zhu, Jason Eshraghian

Abstract: Large language models (LLMs) deliver impressive performance but require large amounts of energy. In this work, we present a MatMul-free LLM architecture adapted for Intel's neuromorphic processor, Loihi 2. Our approach leverages Loihi 2's support for low-precision, event-driven computation and stateful processing. Our hardware-aware quantized model on GPU demonstrates that a 370M parameter MatMul-… ▽ More Large language models (LLMs) deliver impressive performance but require large amounts of energy. In this work, we present a MatMul-free LLM architecture adapted for Intel's neuromorphic processor, Loihi 2. Our approach leverages Loihi 2's support for low-precision, event-driven computation and stateful processing. Our hardware-aware quantized model on GPU demonstrates that a 370M parameter MatMul-free model can be quantized with no accuracy loss. Based on preliminary results, we report up to 3x higher throughput with 2x less energy, compared to transformer-based LLMs on an edge GPU, with significantly better scaling. Further hardware optimizations will increase throughput and decrease energy consumption. These results show the potential of neuromorphic hardware for efficient inference and pave the way for efficient reasoning models capable of generating complex, long-form text rapidly and cost-effectively. △ Less

Submitted 25 March, 2025; v1 submitted 11 February, 2025; originally announced March 2025.

Comments: Accepted to International Conference on Learning Representations (ICLR) Workshop on Scalable Optimization for Efficient and Adaptive Foundation Models (SCOPE)

Accelerating Linear Recurrent Neural Networks for the Edge with Unstructured Sparsity

Authors: Alessandro Pierro, Steven Abreu, Jonathan Timcheck, Philipp Stratmann, Andreas Wild, Sumit Bam Shrestha

Abstract: Linear recurrent neural networks enable powerful long-range sequence modeling with constant memory usage and time-per-token during inference. These architectures hold promise for streaming applications at the edge, but deployment in resource-constrained environments requires hardware-aware optimizations to minimize latency and energy consumption. Unstructured sparsity offers a compelling solution,… ▽ More Linear recurrent neural networks enable powerful long-range sequence modeling with constant memory usage and time-per-token during inference. These architectures hold promise for streaming applications at the edge, but deployment in resource-constrained environments requires hardware-aware optimizations to minimize latency and energy consumption. Unstructured sparsity offers a compelling solution, enabling substantial reductions in compute and memory requirements--when accelerated by compatible hardware platforms. In this paper, we conduct a scaling study to investigate the Pareto front of performance and efficiency across inference compute budgets. We find that highly sparse linear RNNs consistently achieve better efficiency-performance trade-offs than dense baselines, with 2x less compute and 36% less memory at iso-accuracy. Our models achieve state-of-the-art results on a real-time streaming task for audio denoising. By quantizing our sparse models to fixed-point arithmetic and deploying them on the Intel Loihi 2 neuromorphic chip for real-time processing, we translate model compression into tangible gains of 42x lower latency and 149x lower energy consumption compared to a dense model on an edge GPU. Our findings showcase the transformative potential of unstructured sparsity, paving the way for highly efficient recurrent neural networks in real-world, resource-constrained environments. △ Less

Submitted 3 February, 2025; originally announced February 2025.

Comments: Under review

The Two-Loop Lipatov Vertex in QCD

Authors: Samuel Abreu, Giuseppe De Laurentis, Giulio Falcioni, Einan Gardi, Calum Milloy, Leonardo Vernazza

Abstract: High-energy factorization of 2 -> 2 amplitudes in QCD has been recently pushed to the next-to-next-to-leading logarithmic order by determining the three-loop gluon Regge trajectory. This was based on computing multi-Reggeon exchanges using rapidity evolution in the shock-wave formalism, and disentangling between the Regge pole and Regge cut contributions. In the present paper we extend the relevan… ▽ More High-energy factorization of 2 -> 2 amplitudes in QCD has been recently pushed to the next-to-next-to-leading logarithmic order by determining the three-loop gluon Regge trajectory. This was based on computing multi-Reggeon exchanges using rapidity evolution in the shock-wave formalism, and disentangling between the Regge pole and Regge cut contributions. In the present paper we extend the relevant theoretical framework to 2 -> 3 processes, and compute all multi-Reggeon exchanges necessary for extracting the two-loop Reggeon-gluon-Reggeon Lipatov vertex from 2 -> 3 amplitudes. Then, specializing general amplitude methods to multi-Regge kinematics, we derive analytic expressions for non-planar two-loop gg -> ggg, gq -> ggq and qq -> qgq QCD amplitudes in that limit. Matching these to the multi-Reggeon computation, we determine the QCD Lipatov vertex in dimensional regularization at two loops through finite terms. We also determine the one-loop vertex through O(epsilon^4). All results are expressed in a compact form in terms of a basis of single-valued generalised polylogarithms, manifesting target-projectile symmetry and reality properties. Furthermore, our basis of functions is explicitly finite in the soft limit, featuring delicate cancellation of spurious rational poles by transcendental functions. Agreement between all three partonic channels, as well agreement of the maximal weight contributions with the super Yang-Mills Lipatov vertex provide robust checks of the result. △ Less

Submitted 29 December, 2024; originally announced December 2024.

Comments: 107 pages, 9 figures

Regge poles and cuts and the Lipatov vertex

Authors: Samuel Abreu, Giulio Falcioni, Einan Gardi, Calum Milloy, Leonardo Vernazza

Abstract: Scattering amplitudes in the high-energy limit can be described in terms of their singularity structure in the complex angular momentum plane, consisting of Regge poles and cuts. In QCD, gluon Reggeization has long been understood as a manifestation of a Regge pole, but until recently Reggeization violation remained largely obscure. New methods, based on iterative solution of rapidity evolution eq… ▽ More Scattering amplitudes in the high-energy limit can be described in terms of their singularity structure in the complex angular momentum plane, consisting of Regge poles and cuts. In QCD, gluon Reggeization has long been understood as a manifestation of a Regge pole, but until recently Reggeization violation remained largely obscure. New methods, based on iterative solution of rapidity evolution equations, facilitate direct computation of components of the amplitude which are mediated by multi-Reggeon exchange, a manifestation of Regge cuts. Upon disentangling the Regge cut from the pole we are now able to extract the pole parameters from state-of-the-art fixed-order computations (3 loops) and make predictions regarding certain components of the amplitude to higher loop orders. In this talk I review the key ideas which led to this progress, describe where we stand in exploring the structure of 2 -> 2 and 2 -> 3 amplitudes in the (multi-) Regge limit, and comment on the interplay between this research and the study of infrared factorization. △ Less

Submitted 29 December, 2024; originally announced December 2024.

Comments: Talk presented at Loops and Legs in Quantum Field Theory (LL2024), 14-19 April, 2024, Wittenberg, Germany. 17 pages, 5 figures

Journal ref: PoS LL2024 (2024) 085

Planar Six-Point Feynman Integrals for Four-Dimensional Gauge Theories

Authors: Samuel Abreu, Pier Francesco Monni, Ben Page, Johann Usovitsch

Abstract: We compute all planar two-loop six-point Feynman integrals entering scattering observables in massless gauge theories such as QCD. A central result of this paper is the formulation of the differential-equations method under the algebraic constraints stemming from four-dimensional kinematics, which in this case leaves only 8 independent scales. We show that these constraints imply that one must com… ▽ More We compute all planar two-loop six-point Feynman integrals entering scattering observables in massless gauge theories such as QCD. A central result of this paper is the formulation of the differential-equations method under the algebraic constraints stemming from four-dimensional kinematics, which in this case leaves only 8 independent scales. We show that these constraints imply that one must compute topologies with only up to 8 propagators, instead of the expected 9. This leads to the decoupling of entire classes of integrals that do not contribute to scattering amplitudes in four dimensional gauge theories. We construct a pure basis and derive their canonical differential equations, of which we discuss the numerical solution. This work marks an important step towards the calculation of massless $2\to 4$ scattering processes at two loops. △ Less

Submitted 27 December, 2024; originally announced December 2024.

Comments: 32 pages, 2 figures

Report number: CERN-TH-2024-221, HU-EP-24/40-RTG

A generalized picture of colour decoherence in dense QCD media

Authors: Samuel Abreu, Xoán Mayo López, Guilherme Milhano, Alba Soto-Ontoso

Abstract: We revisit the calculation of the soft gluon emission probability off a colour-singlet $q\bar q$ system that evolves in a quark-gluon plasma. The $q\bar q$ antenna is created in the presence of a medium and then emits a soft gluon outside. The gluon emission probability is modified with respect to the vacuum baseline due to interactions with the medium during the formation of the antenna and its p… ▽ More We revisit the calculation of the soft gluon emission probability off a colour-singlet $q\bar q$ system that evolves in a quark-gluon plasma. The $q\bar q$ antenna is created in the presence of a medium and then emits a soft gluon outside. The gluon emission probability is modified with respect to the vacuum baseline due to interactions with the medium during the formation of the antenna and its propagation. Previous studies disregarded the former effect and found that the medium modification to the interference pattern of the antenna was controlled by the so-called critical angle $θ_c$, that exclusively depends on medium properties. We find that accounting for medium interactions during the antenna formation enhances the total rate of emissions off the $q\bar q$ antenna. Interestingly, it also promotes the notion of a critical angle to a dynamic quantity, denoted $\tildeθ_c$, that depends on both the medium and the antenna properties and is thus different for every splitting. As a consequence, depending on the region of parameter space, colour decoherence can either be delayed or accelerated with respect to previous estimates. △ Less

Submitted 31 October, 2024; originally announced October 2024.

Neuromorphic Programming: Emerging Directions for Brain-Inspired Hardware

Authors: Steven Abreu, Jens E. Pedersen

Abstract: The value of brain-inspired neuromorphic computers critically depends on our ability to program them for relevant tasks. Currently, neuromorphic hardware often relies on machine learning methods adapted from deep learning. However, neuromorphic computers have potential far beyond deep learning if we can only harness their energy efficiency and full computational power. Neuromorphic programming wil… ▽ More The value of brain-inspired neuromorphic computers critically depends on our ability to program them for relevant tasks. Currently, neuromorphic hardware often relies on machine learning methods adapted from deep learning. However, neuromorphic computers have potential far beyond deep learning if we can only harness their energy efficiency and full computational power. Neuromorphic programming will necessarily be different from conventional programming, requiring a paradigm shift in how we think about programming. This paper presents a conceptual analysis of programming within the context of neuromorphic computing, challenging conventional paradigms and proposing a framework that aligns more closely with the physical intricacies of these systems. Our analysis revolves around five characteristics that are fundamental to neuromorphic programming and provides a basis for comparison to contemporary programming methods and languages. By studying past approaches, we contribute a framework that advocates for underutilized techniques and calls for richer abstractions to effectively instrument the new hardware class. △ Less

Submitted 15 October, 2024; originally announced October 2024.

Comments: Accepted to International Conference on Neuromorphic Systems (ICONS) 2024. arXiv admin note: substantial text overlap with arXiv:2310.18260

Steering Large Language Models using Conceptors: Improving Addition-Based Activation Engineering

Authors: Joris Postmus, Steven Abreu

Abstract: Large language models have transformed AI, yet reliably controlling their outputs remains a challenge. This paper explores activation engineering, where outputs of pre-trained LLMs are controlled by manipulating their activations at inference time. Unlike traditional methods using a single steering vector, we introduce conceptors - mathematical constructs that represent sets of activation vectors… ▽ More Large language models have transformed AI, yet reliably controlling their outputs remains a challenge. This paper explores activation engineering, where outputs of pre-trained LLMs are controlled by manipulating their activations at inference time. Unlike traditional methods using a single steering vector, we introduce conceptors - mathematical constructs that represent sets of activation vectors as ellipsoidal regions. Conceptors act as soft projection matrices and offer more precise control over complex activation patterns. Our experiments demonstrate that conceptors outperform traditional methods across multiple steering tasks. We further use Boolean operations on conceptors for combined steering goals that empirically outperform additively combining steering vectors on a set of tasks. These results highlight conceptors as a promising tool for more effective steering of LLMs. Our code is available on github.com/jorispos/conceptorsteering. △ Less

Submitted 12 May, 2025; v1 submitted 9 October, 2024; originally announced October 2024.

Comments: Presented at the MINT workshop at NeurIPS 2024. v4: fix sign in equation 10

Strong Electron-Phonon Coupling and Lattice Dynamics in One-Dimensional [(CH3)2NH2]PbI3 Hybrid Perovskite

Authors: A. Nonato, Juan S. Rodríguez-Hernández, D. S. Abreu, C. C. S. Soares, Mayra A. P. Gómez, Alberto García-Fernández, María A. Señarís-Rodríguez, Manuel Sánchez andújar, A. P. Ayala, C. W. A. Paschoal, Rosivaldo Xavier da Silva

Abstract: Hybrid halide perovskites (HHPs) have attracted significant attention due to their remarkable optoelectronic properties that combine the advantages of low cost-effective fabrication methods of organic-inorganic materials. Notably, low-dimensional hybrid halide perovskites including two-dimensional (2D) layers and one-dimensional (1D) chains, are recognized for their superior stability and moisture… ▽ More Hybrid halide perovskites (HHPs) have attracted significant attention due to their remarkable optoelectronic properties that combine the advantages of low cost-effective fabrication methods of organic-inorganic materials. Notably, low-dimensional hybrid halide perovskites including two-dimensional (2D) layers and one-dimensional (1D) chains, are recognized for their superior stability and moisture resistance, making them highly appealing for practical applications. Particularly, DMAPbI3 has attracted attention due to other interesting behaviors and properties, such as thermally induced order-disorder processes, dielectric transition, and cooperative electric ordering of DMA dipole moments. In this paper, we investigated the interplay between low-temperature SPT undergone by the low-dimensional (1D) hybrid halide perovskite-like material DMAPbI3 and its optoelectronic properties. Our approach combines synchrotron X-ray powder diffraction, Raman spectroscopy, thermo-microscopy, differential scanning calorimetry (DSC), and photoluminescence (PL) techniques. Temperature-dependent Synchrotron powder diffraction and Raman Spectroscopy reveal that the modes associated with I-Pb-I and DMA+ ion play a crucial role in the order-disorder SPT in DMAPbI3. The reversible SPT modifies its optoelectronic properties, notably affecting its thermochromic behavior and PL emission. The origin of the PL phenomenon is associated to self-trapped excitons (STEs), which are allowed due to a strong electron-phonon coupling quantified by the Huang-Rhys factor (S = 97+-1). Notably, we identify the longitudinal optical (LO) phonon mode at 84 cm-1 which plays a significant role in electron-phonon interaction. Our results show these STEs not only intensify the PL spectra at lower temperatures but also induce a shift in the color emission, transforming it from a light orange-red to an intense bright strong red. △ Less

Submitted 12 September, 2024; originally announced September 2024.

Comments: 38 pages, 8 figures

EmBARDiment: an Embodied AI Agent for Productivity in XR

Authors: Riccardo Bovo, Steven Abreu, Karan Ahuja, Eric J Gonzalez, Li-Te Cheng, Mar Gonzalez-Franco

Abstract: XR devices running chat-bots powered by Large Language Models (LLMs) have the to become always-on agents that enable much better productivity scenarios. Current screen based chat-bots do not take advantage of the the full-suite of natural inputs available in XR, including inward facing sensor data, instead they over-rely on explicit voice or text prompts, sometimes paired with multi-modal data dro… ▽ More XR devices running chat-bots powered by Large Language Models (LLMs) have the to become always-on agents that enable much better productivity scenarios. Current screen based chat-bots do not take advantage of the the full-suite of natural inputs available in XR, including inward facing sensor data, instead they over-rely on explicit voice or text prompts, sometimes paired with multi-modal data dropped as part of the query. We propose a solution that leverages an attention framework that derives context implicitly from user actions, eye-gaze, and contextual memory within the XR environment. Our work minimizes the need for engineered explicit prompts, fostering grounded and intuitive interactions that glean user insights for the chat-bot. △ Less

Submitted 11 March, 2025; v1 submitted 15 August, 2024; originally announced August 2024.

Journal ref: IEEE Virtual Reality Conference 2025

Two-Loop Five-Point Two-Mass Planar Integrals and Double Lagrangian Insertions in a Wilson Loop

Authors: Samuel Abreu, Dmitry Chicherin, Vasily Sotnikov, Simone Zoia

Abstract: We consider the complete set of planar two-loop five-point Feynman integrals with two off-shell external legs. These integrals are relevant, for instance, for the calculation of the second-order QCD corrections to the production of two heavy vector bosons in association with a jet or a photon at a hadron collider. We construct pure bases for these integrals and reconstruct their analytic different… ▽ More We consider the complete set of planar two-loop five-point Feynman integrals with two off-shell external legs. These integrals are relevant, for instance, for the calculation of the second-order QCD corrections to the production of two heavy vector bosons in association with a jet or a photon at a hadron collider. We construct pure bases for these integrals and reconstruct their analytic differential equations in canonical form through numerical sampling over finite fields. The newly identified symbol alphabet, one of the most complex to date, provides valuable data for bootstrap methods. We then apply our results to initiate the study of double Lagrangian insertions in a four-cusp Wilson loop in planar maximally supersymmetric Yang-Mills theory, computing it through two loops. We observe that it is finite, conformally invariant in four dimensions, and of uniform transcendentality. Furthermore, we provide numerical evidence for its positivity within the amplituhedron region through two loops. △ Less

Submitted 6 November, 2024; v1 submitted 9 August, 2024; originally announced August 2024.

Comments: 33 pages, 6 figures, 2 tables, 2 appendices, ancillary files at https://zenodo.org/records/13254698

Report number: CERN-TH-2024-136, LAPTH-043/24, ZU-TH 40/24

Journal ref: J. High Energ. Phys. 2024, 167 (2024)

Four-loop two-mass tadpoles and the $ρ$ parameter

Authors: Samuel Abreu, Arnd Behring, Andrew McLeod, Ben Page

Abstract: We calculate four-loop QCD corrections to the electroweak $ρ$ parameter with a non-vanishing $b$ quark mass. At three loops, it was observed that elliptic integrals contribute to this observable. This prompts the question of which classes of functions appear at the next order. We report on the status of our calculation with a focus on the mathematical structures that emerge at four loops. We calculate four-loop QCD corrections to the electroweak $ρ$ parameter with a non-vanishing $b$ quark mass. At three loops, it was observed that elliptic integrals contribute to this observable. This prompts the question of which classes of functions appear at the next order. We report on the status of our calculation with a focus on the mathematical structures that emerge at four loops. △ Less

Submitted 31 July, 2024; originally announced July 2024.

Comments: Contribution to the proceedings of the Loops and Legs in Quantum Field Theory conference (LL2024), 14-19 April 2024, Wittenberg, Germany

Report number: CERN-TH-2024-129

Mamba-PTQ: Outlier Channels in Recurrent Large Language Models

Authors: Alessandro Pierro, Steven Abreu

Abstract: Modern recurrent layers are emerging as a promising path toward edge deployment of foundation models, especially in the context of large language models (LLMs). Compressing the whole input sequence in a finite-dimensional representation enables recurrent layers to model long-range dependencies while maintaining a constant inference cost for each token and a fixed memory requirement. However, the p… ▽ More Modern recurrent layers are emerging as a promising path toward edge deployment of foundation models, especially in the context of large language models (LLMs). Compressing the whole input sequence in a finite-dimensional representation enables recurrent layers to model long-range dependencies while maintaining a constant inference cost for each token and a fixed memory requirement. However, the practical deployment of LLMs in resource-limited environments often requires further model compression, such as quantization and pruning. While these techniques are well-established for attention-based models, their effects on recurrent layers remain underexplored. In this preliminary work, we focus on post-training quantization for recurrent LLMs and show that Mamba models exhibit the same pattern of outlier channels observed in attention-based LLMs. We show that the reason for the difficulty of quantizing SSMs is caused by activation outliers, similar to those observed in transformer-based LLMs. We report baseline results for post-training quantization of Mamba that do not take into account the activation outliers and suggest first steps for outlier-aware quantization. △ Less

Submitted 17 July, 2024; originally announced July 2024.

Comments: Work presented at the Efficient Systems for Foundation Models Workshop @ ICML2024

PARSE-Ego4D: Personal Action Recommendation Suggestions for Egocentric Videos

Authors: Steven Abreu, Tiffany D. Do, Karan Ahuja, Eric J. Gonzalez, Lee Payne, Daniel McDuff, Mar Gonzalez-Franco

Abstract: Intelligent assistance involves not only understanding but also action. Existing ego-centric video datasets contain rich annotations of the videos, but not of actions that an intelligent assistant could perform in the moment. To address this gap, we release PARSE-Ego4D, a new set of personal action recommendation annotations for the Ego4D dataset. We take a multi-stage approach to generating and e… ▽ More Intelligent assistance involves not only understanding but also action. Existing ego-centric video datasets contain rich annotations of the videos, but not of actions that an intelligent assistant could perform in the moment. To address this gap, we release PARSE-Ego4D, a new set of personal action recommendation annotations for the Ego4D dataset. We take a multi-stage approach to generating and evaluating these annotations. First, we used a prompt-engineered large language model (LLM) to generate context-aware action suggestions and identified over 18,000 action suggestions. While these synthetic action suggestions are valuable, the inherent limitations of LLMs necessitate human evaluation. To ensure high-quality and user-centered recommendations, we conducted a large-scale human annotation study that provides grounding in human preferences for all of PARSE-Ego4D. We analyze the inter-rater agreement and evaluate subjective preferences of participants. Based on our synthetic dataset and complete human annotations, we propose several new tasks for action suggestions based on ego-centric videos. We encourage novel solutions that improve latency and energy requirements. The annotations in PARSE-Ego4D will support researchers and developers who are working on building action recommendation systems for augmented and virtual reality systems. △ Less

Submitted 25 July, 2024; v1 submitted 14 June, 2024; originally announced July 2024.

Q-S5: Towards Quantized State Space Models

Authors: Steven Abreu, Jens E. Pedersen, Kade M. Heckel, Alessandro Pierro

Abstract: In the quest for next-generation sequence modeling architectures, State Space Models (SSMs) have emerged as a potent alternative to transformers, particularly for their computational efficiency and suitability for dynamical systems. This paper investigates the effect of quantization on the S5 model to understand its impact on model performance and to facilitate its deployment to edge and resource-… ▽ More In the quest for next-generation sequence modeling architectures, State Space Models (SSMs) have emerged as a potent alternative to transformers, particularly for their computational efficiency and suitability for dynamical systems. This paper investigates the effect of quantization on the S5 model to understand its impact on model performance and to facilitate its deployment to edge and resource-constrained platforms. Using quantization-aware training (QAT) and post-training quantization (PTQ), we systematically evaluate the quantization sensitivity of SSMs across different tasks like dynamical systems modeling, Sequential MNIST (sMNIST) and most of the Long Range Arena (LRA). We present fully quantized S5 models whose test accuracy drops less than 1% on sMNIST and most of the LRA. We find that performance on most tasks degrades significantly for recurrent weights below 8-bit precision, but that other components can be compressed further without significant loss of performance. Our results further show that PTQ only performs well on language-based LRA tasks whereas all others require QAT. Our investigation provides necessary insights for the continued development of efficient and hardware-optimized SSMs. △ Less

Submitted 13 June, 2024; originally announced June 2024.

Neuromorphic Intermediate Representation: A Unified Instruction Set for Interoperable Brain-Inspired Computing

Authors: Jens E. Pedersen, Steven Abreu, Matthias Jobst, Gregor Lenz, Vittorio Fra, Felix C. Bauer, Dylan R. Muir, Peng Zhou, Bernhard Vogginger, Kade Heckel, Gianvito Urgese, Sadasivan Shankar, Terrence C. Stewart, Sadique Sheik, Jason K. Eshraghian

Abstract: Spiking neural networks and neuromorphic hardware platforms that simulate neuronal dynamics are getting wide attention and are being applied to many relevant problems using Machine Learning. Despite a well-established mathematical foundation for neural dynamics, there exists numerous software and hardware solutions and stacks whose variability makes it difficult to reproduce findings. Here, we est… ▽ More Spiking neural networks and neuromorphic hardware platforms that simulate neuronal dynamics are getting wide attention and are being applied to many relevant problems using Machine Learning. Despite a well-established mathematical foundation for neural dynamics, there exists numerous software and hardware solutions and stacks whose variability makes it difficult to reproduce findings. Here, we establish a common reference frame for computations in digital neuromorphic systems, titled Neuromorphic Intermediate Representation (NIR). NIR defines a set of computational and composable model primitives as hybrid systems combining continuous-time dynamics and discrete events. By abstracting away assumptions around discretization and hardware constraints, NIR faithfully captures the computational model, while bridging differences between the evaluated implementation and the underlying mathematical formalism. NIR supports an unprecedented number of neuromorphic systems, which we demonstrate by reproducing three spiking neural network models of different complexity across 7 neuromorphic simulators and 4 digital hardware platforms. NIR decouples the development of neuromorphic hardware and software, enabling interoperability between platforms and improving accessibility to multiple neuromorphic technologies. We believe that NIR is a key next step in brain-inspired hardware-software co-evolution, enabling research towards the implementation of energy efficient computational principles of nervous systems. NIR is available at neuroir.org △ Less

Submitted 30 September, 2024; v1 submitted 24 November, 2023; originally announced November 2023.

Comments: NIR is available at https://neuroir.org

Journal ref: Nat Commun 15, 8122 (2024)

Concepts and Paradigms for Neuromorphic Programming

Authors: Steven Abreu

Abstract: The value of neuromorphic computers depends crucially on our ability to program them for relevant tasks. Currently, neuromorphic computers are mostly limited to machine learning methods adapted from deep learning. However, neuromorphic computers have potential far beyond deep learning if we can only make use of their computational properties to harness their full power. Neuromorphic programming wi… ▽ More The value of neuromorphic computers depends crucially on our ability to program them for relevant tasks. Currently, neuromorphic computers are mostly limited to machine learning methods adapted from deep learning. However, neuromorphic computers have potential far beyond deep learning if we can only make use of their computational properties to harness their full power. Neuromorphic programming will necessarily be different from conventional programming, requiring a paradigm shift in how we think about programming in general. The contributions of this paper are 1) a conceptual analysis of what "programming" means in the context of neuromorphic computers and 2) an exploration of existing programming paradigms that are promising yet overlooked in neuromorphic computing. The goal is to expand the horizon of neuromorphic programming methods, thereby allowing researchers to move beyond the shackles of current methods and explore novel directions. △ Less

Submitted 27 October, 2023; originally announced October 2023.

Demonstrating (Hybrid) Active Logic Documents and the Ciao Prolog Playground, and an Application to Verification Tutorials

Authors: Daniela Ferreiro, José F. Morales, Salvador Abreu, Manuel V. Hermenegildo

Abstract: Active Logic Documents (ALD) are web pages which incorporate embedded Prolog engines that run locally within the browser. ALD offers both a very easy way to add click-to-run capabilities to any kind of teaching materials, independently of the tool used to generate them, as well as a tool-set for generating web-based materials with embedded examples and exercises. Both leverage on (components of)… ▽ More Active Logic Documents (ALD) are web pages which incorporate embedded Prolog engines that run locally within the browser. ALD offers both a very easy way to add click-to-run capabilities to any kind of teaching materials, independently of the tool used to generate them, as well as a tool-set for generating web-based materials with embedded examples and exercises. Both leverage on (components of) the Ciao Prolog Playground. We present a demonstration of the ALD approach and the Ciao Prolog Playground, as well as a recent extension to ALDs to facilitate the integration of other tools into the system for creating Hybrid Active Logic Documents (HALD). We also present a concrete application of these technologies to the creation of tutorials for a program verification tool. △ Less

Submitted 30 August, 2023; originally announced August 2023.

Comments: In Proceedings ICLP 2023, arXiv:2308.14898

Journal ref: EPTCS 385, 2023, pp. 324-329

All Two-Loop Feynman Integrals for Five-Point One-Mass Scattering

Authors: Samuel Abreu, Dmitry Chicherin, Harald Ita, Ben Page, Vasily Sotnikov, Wladimir Tschernow, Simone Zoia

Abstract: We compute the complete set of two-loop master integrals for the scattering of four massless particles and a massive one. Our results are ready for phenomenological applications, removing a major obstacle to the computation of complete next-to-next-to-leading order (NNLO) QCD corrections to processes such as the production of a $H/Z/W$ boson in association with two jets at the LHC. Furthermore, th… ▽ More We compute the complete set of two-loop master integrals for the scattering of four massless particles and a massive one. Our results are ready for phenomenological applications, removing a major obstacle to the computation of complete next-to-next-to-leading order (NNLO) QCD corrections to processes such as the production of a $H/Z/W$ boson in association with two jets at the LHC. Furthermore, they open the door to new investigations into the structure of quantum-field theories and provide precious analytic data for studying the mathematical properties of Feynman integrals. △ Less

Submitted 2 May, 2024; v1 submitted 27 June, 2023; originally announced June 2023.

Comments: Ancillary files can be downloaded from https://doi.org/10.5281/zenodo.8085740 ; v2 (published version): added appendix describing the analytic continuation across logarithmic singularities within the physical region

Two-Loop QCD Corrections for Three-Photon Production at Hadron Colliders

Authors: Samuel Abreu, Giuseppe De Laurentis, Harald Ita, Maximillian Klinkert, Ben Page, Vasily Sotnikov

Abstract: We complete the computation of the two-loop helicity amplitudes for the production of three photons at hadron colliders, including all contributions beyond the leading-color approximation. We reconstruct the analytic form of the amplitudes from numerical finite-field samples obtained with the numerical unitarity method. This method requires as input surface terms for all relevant five-point non-pl… ▽ More We complete the computation of the two-loop helicity amplitudes for the production of three photons at hadron colliders, including all contributions beyond the leading-color approximation. We reconstruct the analytic form of the amplitudes from numerical finite-field samples obtained with the numerical unitarity method. This method requires as input surface terms for all relevant five-point non-planar integral topologies, which we obtain by solving the associated syzygy problem in embedding space. The numerical samples are used to constrain compact spinor-helicity ansätze, which are optimized by taking advantage of the known one-loop analytic structure. We make our analytic results available in a public C++ library, which is suitable for immediate phenomenological applications. We estimate that the inclusion of the subleading-color contributions will decrease the size of the two-loop corrections by about 30% to 50%, and the NNLO cross sections by a few percent, compared to the results in the leading-color approximation. △ Less

Submitted 13 October, 2023; v1 submitted 26 May, 2023; originally announced May 2023.

Comments: 38 pages, 5 figures, 6 tables - published version

Report number: CERN-TH-2023-090, PSI-PR-23-16, ZU-TH 23/23

Journal ref: SciPost Phys. 15, 157 (2023)

Training a spiking neural network on an event-based label-free flow cytometry dataset

Authors: Muhammed Gouda, Steven Abreu, Alessio Lugnan, Peter Bienstman

Abstract: Imaging flow cytometry systems aim to analyze a huge number of cells or micro-particles based on their physical characteristics. The vast majority of current systems acquire a large amount of images which are used to train deep artificial neural networks. However, this approach increases both the latency and power consumption of the final apparatus. In this work-in-progress, we combine an event-ba… ▽ More Imaging flow cytometry systems aim to analyze a huge number of cells or micro-particles based on their physical characteristics. The vast majority of current systems acquire a large amount of images which are used to train deep artificial neural networks. However, this approach increases both the latency and power consumption of the final apparatus. In this work-in-progress, we combine an event-based camera with a free-space optical setup to obtain spikes for each particle passing in a microfluidic channel. A spiking neural network is trained on the collected dataset, resulting in 97.7% mean training accuracy and 93.5% mean testing accuracy for the fully event-based classification pipeline. △ Less

Submitted 19 March, 2023; originally announced March 2023.

Comments: Accepted to Neuro-Inspired Computational Elements (NICE) conference by ACM in San Antonio, TX, USA, 2023

Two-loop form factors for pseudo-scalar quarkonium production and decay

Authors: Samuel Abreu, Matteo Becchetti, Claude Duhr, Melih A. Ozcelik

Abstract: We present the analytic expressions for the two-loop form factors for the production or decay of pseudo-scalar quarkonia, in a scheme where the quarks are produced at threshold. We consider the two-loop amplitude for the process $γγ\leftrightarrow {^1S_0^{[1]}}$, that was previously known only numerically, as well as for the processes $gg \leftrightarrow {^1S_0^{[1]}}$,… ▽ More We present the analytic expressions for the two-loop form factors for the production or decay of pseudo-scalar quarkonia, in a scheme where the quarks are produced at threshold. We consider the two-loop amplitude for the process $γγ\leftrightarrow {^1S_0^{[1]}}$, that was previously known only numerically, as well as for the processes $gg \leftrightarrow {^1S_0^{[1]}}$, $γg \leftrightarrow {^1S_0^{[8]}}$ and $gg \leftrightarrow {^1S_0^{[8]}}$, which have not been computed before. The two-loop corrections to $gg \leftrightarrow {^1S_0^{[1]}}$ are the last missing ingredients for a full NNLO calculation of $η_Q$ hadro-production. We discuss how the singularity structure of the amplitudes is affected by the threshold kinematics, which in particular introduces Coulomb singularities. In this context, we first show how the usual structure of the infrared singularities degenerates at threshold kinematics, and then extract the anomalous dimensions governing the Coulomb singularities for colour-singlet and octet channels, the latter being presented here for the first time. We give high-precision numerical results for the hard functions, which can be used for phenomenological studies of $η_Q$ production and decay at NNLO. △ Less

Submitted 1 March, 2023; v1 submitted 16 November, 2022; originally announced November 2022.

Comments: 37 pages, 1 figure, v2: version published in JHEP

Report number: BONN-TH-2022-23, CERN-TH-2022-187, TTP22-068

The Diagrammatic Coaction

Authors: Samuel Abreu, Ruth Britto, Claude Duhr, Einan Gardi, James Matthew

Abstract: The diagrammatic coaction underpins the analytic structure of Feynman integrals, their cuts and the differential equations they admit. The coaction maps any diagram into a tensor product of its pinches and cuts. These correspond respectively to differential forms defining master integrals, and integration contours which place a subset of the propagators on shell. In a canonical basis these forms a… ▽ More The diagrammatic coaction underpins the analytic structure of Feynman integrals, their cuts and the differential equations they admit. The coaction maps any diagram into a tensor product of its pinches and cuts. These correspond respectively to differential forms defining master integrals, and integration contours which place a subset of the propagators on shell. In a canonical basis these forms and contours are dual to each other. In this talk I review our present understanding of this algebraic structure and its manifestation for dimensionally-regularized Feynman integrals that are expandable to polylogarithms around integer dimensions. Using one- and two-loop integral examples, I will explain the duality between forms and contours, and the correspondence between the local coaction acting on the Laurent coefficients in the dimensional regulator and the global coaction acting on generalised hypergeometric functions. △ Less

Submitted 16 July, 2022; originally announced July 2022.

Comments: 19 pages, Talk presented at Loop and Legs in Quantum Field Theory - LL2022, 25-30 April, 2022, Ettal, Germany

Report number: CERN-TH-2022-122, BONN-TH-2022-18

Quark and gluon two-loop beam functions for leading-jet $p_T$ and slicing at NNLO

Authors: Samuel Abreu, Jonathan R. Gaunt, Pier Francesco Monni, Luca Rottoli, Robert Szafron

Abstract: We compute the complete set of two-loop beam functions for the transverse momentum distribution of the leading jet produced in association with an arbitrary colour-singlet system. Our results constitute the last missing ingredient for the calculation of the jet-vetoed cross section at small veto scales at the next-to-next-to-leading order, as well as an important ingredient for its resummation to… ▽ More We compute the complete set of two-loop beam functions for the transverse momentum distribution of the leading jet produced in association with an arbitrary colour-singlet system. Our results constitute the last missing ingredient for the calculation of the jet-vetoed cross section at small veto scales at the next-to-next-to-leading order, as well as an important ingredient for its resummation to next-to-next-to-next-to-leading logarithmic order. Our calculation is performed in the soft-collinear effective theory framework with a suitable regularisation of the rapidity divergences occurring in the phase-space integrals. We discuss the occurrence of soft-collinear mixing terms that might violate the factorisation theorem, and demonstrate that they vanish at two loops in the exponential rapidity regularisation scheme when performing a multipole expansion of the measurement function. As in our recent computation of the two-loop soft function, we present the results as a Laurent expansion in the jet radius $R$. We provide analytic expressions for all flavour channels in $x$ space with the exception of a set of $R$-independent non-logarithmic terms that are given as numerical grids. We also perform a fully numerical calculation with exact $R$ dependence, and find that it agrees with our analytic expansion at the permyriad level or better. Our calculation allows us to define a next-to-next-to-leading order slicing method using the leading-jet $p_T$ as a slicing variable. As a check of our results, we carry out a calculation of the Higgs and $Z$ boson total production cross sections at the next-to-next-to-leading order in QCD. △ Less

Submitted 14 July, 2022; originally announced July 2022.

Comments: 26 pages, 3 figures and a set of ancillary files

Report number: CERN-TH-2022-118, ZU-TH 30/22

Two-loop master integrals for pseudo-scalar quarkonium and leptonium production and decay

Authors: Samuel Abreu, Matteo Becchetti, Claude Duhr, Melih A. Ozcelik

Abstract: We compute the master integrals relevant for the two-loop corrections to pseudo-scalar quarkonium and leptonium production and decay. We present both analytic and high-precision numerical results. The analytic expressions are given in terms of multiple polylogarithms (MPLs), elliptic multiple polylogarithms (eMPLs) and iterated integrals of Eisenstein series. As an application of our results, we o… ▽ More We compute the master integrals relevant for the two-loop corrections to pseudo-scalar quarkonium and leptonium production and decay. We present both analytic and high-precision numerical results. The analytic expressions are given in terms of multiple polylogarithms (MPLs), elliptic multiple polylogarithms (eMPLs) and iterated integrals of Eisenstein series. As an application of our results, we obtain for the first time an analytic expression for the two-loop amplitude for para-positronium decay to two photons at two loops. △ Less

Submitted 17 November, 2022; v1 submitted 8 June, 2022; originally announced June 2022.

Comments: 34 pages, 6 figures, v2: references updated, version published in JHEP

Report number: BONN-TH-2022-14, CERN-TH-2022-093, TTP22-037

The analytic two-loop soft function for leading-jet $p_T$

Authors: Samuel Abreu, Jonathan R. Gaunt, Pier Francesco Monni, Robert Szafron

Abstract: We present the calculation of the two-loop soft function for the transverse momentum distribution of the leading jet produced in association with any colour-singlet system (e.g.~a Higgs or a $Z$ boson). This constitutes a central ingredient for the resummation of the above distribution as well as the jet-vetoed cross section at the next-to-next-to-next-to-leading logarithmic order, both of which p… ▽ More We present the calculation of the two-loop soft function for the transverse momentum distribution of the leading jet produced in association with any colour-singlet system (e.g.~a Higgs or a $Z$ boson). This constitutes a central ingredient for the resummation of the above distribution as well as the jet-vetoed cross section at the next-to-next-to-next-to-leading logarithmic order, both of which play an important role in the precision physics programme at the Large Hadron Collider. The calculation is performed in soft-collinear effective theory with an appropriate regularisation of the rapidity divergences that occur in the phase-space integrals. We obtain analytic results by employing an exponential regulator and by taking a Laurent expansion in the jet radius $R$. All expressions are presented as ancillary files with this article. △ Less

Submitted 6 April, 2022; originally announced April 2022.

Comments: 17 pages, 4 figures, and ancillary files with results

Report number: CERN-TH-2022-059

The SAGEX Review on Scattering Amplitudes, Chapter 3: Mathematical structures in Feynman integrals

Authors: Samuel Abreu, Ruth Britto, Claude Duhr

Abstract: Dimensionally-regulated Feynman integrals are a cornerstone of all perturbative computations in quantum field theory. They are known to exhibit a rich mathematical structure, which has led to the development of powerful new techniques for their computation. We review some of the most recent advances in our understanding of the analytic structure of multiloop Feynman integrals in dimensional regula… ▽ More Dimensionally-regulated Feynman integrals are a cornerstone of all perturbative computations in quantum field theory. They are known to exhibit a rich mathematical structure, which has led to the development of powerful new techniques for their computation. We review some of the most recent advances in our understanding of the analytic structure of multiloop Feynman integrals in dimensional regularisation. In particular, we give an overview of modern approaches to computing Feynman integrals using differential equations, and we discuss some of the properties of the functions that appear in the solutions. We then review how dimensional regularisation has a natural mathematical interpretation in terms of the theory of twisted cohomology groups, and how many of the well-known ideas about Feynman integrals arise naturally in this context. This is Chapter 3 of a series of review articles on scattering amplitudes, of which Chapter 0 [arXiv:2203.13011] presents an overview and Chapter 4 [arXiv:2203.13015] contains closely related topics. △ Less

Submitted 10 January, 2023; v1 submitted 24 March, 2022; originally announced March 2022.

Comments: 62 pages, see also the overview article arXiv:2203.13011. v3: journal version

Report number: SAGEX-22-04, BONN-TH-2022-03, CERN-TH-2022-021

Journal ref: J.Phys.A 55 (2022) 44, 443004

The SAGEX Review on Scattering Amplitudes

Authors: Gabriele Travaglini, Andreas Brandhuber, Patrick Dorey, Tristan McLoughlin, Samuel Abreu, Zvi Bern, N. Emil J. Bjerrum-Bohr, Johannes Blümlein, Ruth Britto, John Joseph M. Carrasco, Dmitry Chicherin, Marco Chiodaroli, Poul H. Damgaard, Vittorio Del Duca, Lance J. Dixon, Daniele Dorigoni, Claude Duhr, Yvonne Geyer, Michael B. Green, Enrico Herrmann, Paul Heslop, Henrik Johansson, Gregory P. Korchemsky, David A. Kosower, Lionel Mason , et al. (13 additional authors not shown)

Abstract: This is an introduction to, and invitation to read, a series of review articles on scattering amplitudes in gauge theory, gravity, and superstring theory. Our aim is to provide an overview of the field, from basic aspects to a selection of current (2022) research and developments. This is an introduction to, and invitation to read, a series of review articles on scattering amplitudes in gauge theory, gravity, and superstring theory. Our aim is to provide an overview of the field, from basic aspects to a selection of current (2022) research and developments. △ Less

Submitted 8 January, 2023; v1 submitted 24 March, 2022; originally announced March 2022.

Comments: 15 pages, overview article. v3: journal version

Report number: SAGEX-22-01

Fifty Years of Prolog and Beyond

Authors: Philipp Körner, Michael Leuschel, João Barbosa, Vítor Santos Costa, Verónica Dahl, Manuel V. Hermenegildo, Jose F. Morales, Jan Wielemaker, Daniel Diaz, Salvador Abreu, Giovanni Ciatto

Abstract: Both logic programming in general, and Prolog in particular, have a long and fascinating history, intermingled with that of many disciplines they inherited from or catalyzed. A large body of research has been gathered over the last 50 years, supported by many Prolog implementations. Many implementations are still actively developed, while new ones keep appearing. Often, the features added by diffe… ▽ More Both logic programming in general, and Prolog in particular, have a long and fascinating history, intermingled with that of many disciplines they inherited from or catalyzed. A large body of research has been gathered over the last 50 years, supported by many Prolog implementations. Many implementations are still actively developed, while new ones keep appearing. Often, the features added by different systems were motivated by the interdisciplinary needs of programmers and implementors, yielding systems that, while sharing the "classic" core language, and, in particular, the main aspects of the ISO-Prolog standard, also depart from each other in other aspects. This obviously poses challenges for code portability. The field has also inspired many related, but quite different languages that have created their own communities. This article aims at integrating and applying the main lessons learned in the process of evolution of Prolog. It is structured into three major parts. Firstly, we overview the evolution of Prolog systems and the community approximately up to the ISO standard, considering both the main historic developments and the motivations behind several Prolog implementations, as well as other logic programming languages influenced by Prolog. Then, we discuss the Prolog implementations that are most active after the appearance of the standard: their visions, goals, commonalities, and incompatibilities. Finally, we perform a SWOT analysis in order to better identify the potential of Prolog, and propose future directions along which Prolog might continue to add useful features, interfaces, libraries, and tools, while at the same time improving compatibility between implementations. △ Less

Submitted 14 March, 2022; v1 submitted 26 January, 2022; originally announced January 2022.

Comments: 87 pages, 2 figures. This article has been accepted for publication in Theory and Practice of Logic Programming (TPLP)

Leading-Color Two-Loop Amplitudes for Four Partons and a W Boson in QCD

Authors: S. Abreu, F. Febres Cordero, H. Ita, M. Klinkert, B. Page, V. Sotnikov

Abstract: We present the leading-color two-loop QCD corrections for the scattering of four partons and a $W$ boson, including its leptonic decay. The amplitudes are assembled from the planar two-loop helicity amplitudes for four partons and a vector boson decaying to a lepton pair, which are also used to determine the planar two-loop amplitudes for four partons and a $Z/γ^*$ boson with a leptonic decay. The… ▽ More We present the leading-color two-loop QCD corrections for the scattering of four partons and a $W$ boson, including its leptonic decay. The amplitudes are assembled from the planar two-loop helicity amplitudes for four partons and a vector boson decaying to a lepton pair, which are also used to determine the planar two-loop amplitudes for four partons and a $Z/γ^*$ boson with a leptonic decay. The analytic expressions are obtained by setting up a dedicated Ansatz and constraining the free parameters from numerical samples obtained within the framework of numerical unitarity. The large linear systems that must be solved to determine the analytic expressions are constructed to be in Vandermonde form. Such systems can be very efficiently solved, bypassing the bottleneck of Gaussian elimination. Our results are expressed in a basis of one-mass pentagon functions, which opens the possibility of their efficient numerical evaluation. △ Less

Submitted 9 April, 2022; v1 submitted 14 October, 2021; originally announced October 2021.

Comments: 31 pages, 5 figures; v3: matches the published version; analytic expressions at http://www.hep.fsu.edu/~ffebres/W4partons

Report number: CERN-TH-2021-156, FR-PHENO-2021-12, MPP-2021-181

Journal ref: J. High Energ. Phys. 2022, 42 (2022)

Two-Loop Hexa-Box Integrals for Non-Planar Five-Point One-Mass Processes

Authors: Samuel Abreu, Harald Ita, Ben Page, Wladimir Tschernow

Abstract: We present the calculation of the three distinct non-planar hexa-box topologies for five-point one-mass processes. These three topologies are required to obtain the two-loop virtual QCD corrections for two-jet-associated W, Z or Higgs-boson production. Each topology is solved by obtaining a pure basis of master integrals and efficiently constructing the associated differential equation with numeri… ▽ More We present the calculation of the three distinct non-planar hexa-box topologies for five-point one-mass processes. These three topologies are required to obtain the two-loop virtual QCD corrections for two-jet-associated W, Z or Higgs-boson production. Each topology is solved by obtaining a pure basis of master integrals and efficiently constructing the associated differential equation with numerical sampling and unitarity-cut techniques. We present compact expressions for the alphabet of these non-planar integrals, and discuss some properties of their symbol. Notably, we observe that the extended Steinmann relations are in general not satisfied. Finally, we solve the differential equations in terms of generalized power series and provide high-precision values in different regions of phase space which can be used as boundary conditions for subsequent evaluations. △ Less

Submitted 29 July, 2021; originally announced July 2021.

The diagrammatic coaction beyond one loop

Authors: Samuel Abreu, Ruth Britto, Claude Duhr, Einan Gardi, James Matthew

Abstract: The diagrammatic coaction maps any given Feynman graph into pairs of graphs and cut graphs such that, conjecturally, when these graphs are replaced by the corresponding Feynman integrals one obtains a coaction on the respective functions. The coaction on the functions is constructed by pairing a basis of differential forms, corresponding to master integrals, with a basis of integration contours, c… ▽ More The diagrammatic coaction maps any given Feynman graph into pairs of graphs and cut graphs such that, conjecturally, when these graphs are replaced by the corresponding Feynman integrals one obtains a coaction on the respective functions. The coaction on the functions is constructed by pairing a basis of differential forms, corresponding to master integrals, with a basis of integration contours, corresponding to independent cut integrals. At one loop, a general diagrammatic coaction was established using dimensional regularisation, which may be realised in terms of a global coaction on hypergeometric functions, or equivalently, order by order in the $ε$ expansion, via a local coaction on multiple polylogarithms. The present paper takes the first steps in generalising the diagrammatic coaction beyond one loop. We first establish general properties that govern the diagrammatic coaction at any loop order. We then focus on examples of two-loop topologies for which all integrals expand into polylogarithms. In each case we determine bases of master integrals and cuts in terms of hypergeometric functions, and then use the global coaction to establish the diagrammatic coaction of all master integrals in the topology. The diagrammatic coaction encodes the complete set of discontinuities of Feynman integrals, as well as the differential equations they satisfy, providing a general tool to understand their physical and mathematical properties. △ Less

Submitted 2 June, 2021; originally announced June 2021.

Leading-Color Two-Loop QCD Corrections for Three-Jet Production at Hadron Colliders

Authors: S. Abreu, F. Febres Cordero, H. Ita, B. Page, V. Sotnikov

Abstract: We present the complete set of leading-color two-loop contributions required to obtain next-to-next-to-leading-order (NNLO) QCD corrections to three-jet production at hadron colliders. We obtain analytic expressions for a generating set of finite remainders, valid in the physical region for three-jet production. The analytic continuation of the known Euclidean-region results is determined from a s… ▽ More We present the complete set of leading-color two-loop contributions required to obtain next-to-next-to-leading-order (NNLO) QCD corrections to three-jet production at hadron colliders. We obtain analytic expressions for a generating set of finite remainders, valid in the physical region for three-jet production. The analytic continuation of the known Euclidean-region results is determined from a small set of numerical evaluations of the amplitudes. We obtain analytic expressions that are suitable for phenomenological applications and we present a C++ library for their efficient and stable numerical evaluation. △ Less

Submitted 10 January, 2024; v1 submitted 26 February, 2021; originally announced February 2021.

Comments: Analytic results in ancillary files; link for numerical code in references. v3: matches the published version. v4: fix notation in eq. (3.12) to match ancillary files

Report number: CERN-TH-2021-023, FR-PHENO-2021-06, MPP-2021-20

Journal ref: J. High Energ. Phys. 2021, 95 (2021)

Leading-Color Two-Loop QCD Corrections for Three-Photon Production at Hadron Colliders

Authors: S. Abreu, B. Page, E. Pascual, V. Sotnikov

Abstract: We compute the two-loop helicity amplitudes for the production of three photons at hadron colliders in QCD at leading-color. Using the two-loop numerical unitarity method coupled with analytic reconstruction techniques, we obtain the decomposition of the two-loop amplitudes in terms of master integrals in analytic form. These expressions are valid to all orders in the dimensional regulator. We use… ▽ More We compute the two-loop helicity amplitudes for the production of three photons at hadron colliders in QCD at leading-color. Using the two-loop numerical unitarity method coupled with analytic reconstruction techniques, we obtain the decomposition of the two-loop amplitudes in terms of master integrals in analytic form. These expressions are valid to all orders in the dimensional regulator. We use them to compute the two-loop finite remainders, which are given in a form that can be efficiently evaluated across the whole physical phase space. We further package these results in a public code which assembles the helicity-summed squared two-loop remainders, whose numerical stability across phase-space is demonstrated. This is the first time that a five-point two-loop process is publicly available for immediate phenomenological applications. △ Less

Submitted 18 January, 2021; v1 submitted 29 October, 2020; originally announced October 2020.

Comments: Minor text changes, updated references, and a typo fix. Matches the published version

Report number: CERN-TH-2020-176, FR-PHENO-2020-016, MPP-2020-190

Journal ref: J. High Energ. Phys. 2021, 78 (2021)

Caravel: A C++ Framework for the Computation of Multi-Loop Amplitudes with Numerical Unitarity

Authors: S. Abreu, J. Dormans, F. Febres Cordero, H. Ita, M. Kraus, B. Page, E. Pascual, M. S. Ruf, V. Sotnikov

Abstract: We present the first public version of Caravel, a C++17 framework for the computation of multi-loop scattering amplitudes in quantum field theory, based on the numerical unitarity method. Caravel is composed of modules for the $D$-dimensional decomposition of integrands of scattering amplitudes into master and surface terms, the computation of tree-level amplitudes in floating point or finite-fiel… ▽ More We present the first public version of Caravel, a C++17 framework for the computation of multi-loop scattering amplitudes in quantum field theory, based on the numerical unitarity method. Caravel is composed of modules for the $D$-dimensional decomposition of integrands of scattering amplitudes into master and surface terms, the computation of tree-level amplitudes in floating point or finite-field arithmetic, the numerical computation of one- and two-loop amplitudes in QCD and Einstein gravity, and functional reconstruction tools. We provide programs that showcase Caravel's main functionalities and allow to compute selected one- and two-loop amplitudes. △ Less

Submitted 24 September, 2020; originally announced September 2020.

Two-Loop Integrals for Planar Five-Point One-Mass Processes

Authors: S. Abreu, H. Ita, F. Moriello, B. Page, W. Tschernow, M. Zeng

Abstract: We present the computation of a full set of planar five-point two-loop master integrals with one external mass. These integrals are an important ingredient for two-loop scattering amplitudes for two-jet-associated W-boson production at leading color in QCD. We provide a set of pure integrals together with differential equations in canonical form. We obtain analytic differential equations efficient… ▽ More We present the computation of a full set of planar five-point two-loop master integrals with one external mass. These integrals are an important ingredient for two-loop scattering amplitudes for two-jet-associated W-boson production at leading color in QCD. We provide a set of pure integrals together with differential equations in canonical form. We obtain analytic differential equations efficiently from numerical samples over finite fields, fitting an ansatz built from symbol letters. The symbol alphabet itself is constructed from cut differential equations and we find that it can be written in a remarkably compact form. We comment on the analytic properties of the integrals and confirm the extended Steinmann relations, which govern the double discontinuities of Feynman integrals, to all orders in $ε$. We solve the differential equations in terms of generalized power series on single-parameter contours in the space of Mandelstam invariants. This form of the solution trivializes the analytic continuation and the integrals can be evaluated in all kinematic regions with arbitrary numerical precision. △ Less

Submitted 27 May, 2020; v1 submitted 8 May, 2020; originally announced May 2020.

Comments: v2: minor typos corrected

The Two-Loop Four-Graviton Scattering Amplitudes

Authors: S. Abreu, F. Febres Cordero, H. Ita, M. Jaquier, B. Page, M. S. Ruf, V. Sotnikov

Abstract: We present the analytic form of the two-loop four-graviton scattering amplitudes in Einstein gravity. To remove ultraviolet divergences we include counterterms quadratic and cubic in the Riemann curvature tensor. The two-loop numerical unitarity approach is used to deal with the challenging momentum dependence of the interactions. We exploit the algebraic properties of the integrand of the amplitu… ▽ More We present the analytic form of the two-loop four-graviton scattering amplitudes in Einstein gravity. To remove ultraviolet divergences we include counterterms quadratic and cubic in the Riemann curvature tensor. The two-loop numerical unitarity approach is used to deal with the challenging momentum dependence of the interactions. We exploit the algebraic properties of the integrand of the amplitude in order to map it to a minimal basis of Feynman integrals. Analytic expressions are obtained from numerical evaluations of the amplitude. Finally, we show that four-graviton scattering observables depend on fewer couplings than naively expected. △ Less

Submitted 27 February, 2020; originally announced February 2020.

Comments: 5 pages, 1 figure

Report number: CP3-20-11, FR-PHENO-2020-002, IPhT-20/003, TTP20-003, MPP-2020-17

Diagrammatic Coaction of Two-Loop Feynman Integrals

Authors: Samuel Abreu, Ruth Britto, Claude Duhr, Einan Gardi, James Matthew

Abstract: It is known that one-loop Feynman integrals possess an algebraic structure encoding some of their analytic properties called the coaction, which can be written in terms of Feynman integrals and their cuts. This diagrammatic coaction, and the coaction on other classes of integrals such as hypergeometric functions, may be expressed using suitable bases of differential forms and integration contours.… ▽ More It is known that one-loop Feynman integrals possess an algebraic structure encoding some of their analytic properties called the coaction, which can be written in terms of Feynman integrals and their cuts. This diagrammatic coaction, and the coaction on other classes of integrals such as hypergeometric functions, may be expressed using suitable bases of differential forms and integration contours. This provides a useful framework for computing coactions of Feynman integrals expressed using the hypergeometric functions. We will discuss examples where this technique has been used in the calculation of two-loop diagrammatic coactions. △ Less

Submitted 13 December, 2019; originally announced December 2019.

Comments: 10 pages, talk given at RADCOR 2019

Report number: CERN-TH-2019-218, CP3-19-59

Generalized hypergeometric functions and intersection theory for Feynman integrals

Authors: Samuel Abreu, Ruth Britto, Claude Duhr, Einan Gardi, James Matthew

Abstract: Feynman integrals that have been evaluated in dimensional regularization can be written in terms of generalized hypergeometric functions. It is well known that properties of these functions are revealed in the framework of intersection theory. We propose a new application of intersection theory to construct a coaction on generalized hypergeometric functions. When applied to dimensionally regulariz… ▽ More Feynman integrals that have been evaluated in dimensional regularization can be written in terms of generalized hypergeometric functions. It is well known that properties of these functions are revealed in the framework of intersection theory. We propose a new application of intersection theory to construct a coaction on generalized hypergeometric functions. When applied to dimensionally regularized Feynman integrals, this coaction reproduces the coaction on multiple polylogarithms order by order in the parameter of dimensional regularization. △ Less

Submitted 10 December, 2019; v1 submitted 6 December, 2019; originally announced December 2019.

Comments: 10 pages, talk given at RADCOR 2019, based on arXiv:1910.08358. v2: F3 coaction formula fixed

Report number: CERN-TH-2019-219, CP3-19-58

Three-loop contributions to the $ρ$ parameter and iterated integrals of modular forms

Authors: Samuel Abreu, Matteo Becchetti, Claude Duhr, Robin Marzucca

Abstract: We compute fully analytic results for the three-loop diagrams involving two different massive quark flavours contributing to the $ρ$ parameter in the Standard Model. We find that the results involve exactly the same class of functions that appears in the well-known sunrise and banana graphs, namely elliptic polylogarithms and iterated integrals of modular forms. Using recent developments in the un… ▽ More We compute fully analytic results for the three-loop diagrams involving two different massive quark flavours contributing to the $ρ$ parameter in the Standard Model. We find that the results involve exactly the same class of functions that appears in the well-known sunrise and banana graphs, namely elliptic polylogarithms and iterated integrals of modular forms. Using recent developments in the understanding of these functions, we analytically continue all the iterated integrals of modular forms to all regions of the parameter space, and in each region we obtain manifestly real and fast-converging series expansions for these functions. △ Less

Submitted 5 December, 2019; originally announced December 2019.

Comments: 37 pages, 2 figures

From positive geometries to a coaction on hypergeometric functions

Authors: Samuel Abreu, Ruth Britto, Claude Duhr, Einan Gardi, James Matthew

Abstract: It is well known that Feynman integrals in dimensional regularization often evaluate to functions of hypergeometric type. Inspired by a recent proposal for a coaction on one-loop Feynman integrals in dimensional regularization, we use intersection numbers and twisted homology theory to define a coaction on certain hypergeometric functions. The functions we consider admit an integral representation… ▽ More It is well known that Feynman integrals in dimensional regularization often evaluate to functions of hypergeometric type. Inspired by a recent proposal for a coaction on one-loop Feynman integrals in dimensional regularization, we use intersection numbers and twisted homology theory to define a coaction on certain hypergeometric functions. The functions we consider admit an integral representation where both the integrand and the contour of integration are associated with positive geometries. As in dimensionally-regularized Feynman integrals, endpoint singularities are regularized by means of exponents controlled by a small parameter $ε$. We show that the coaction defined on this class of integral is consistent, upon expansion in $ε$, with the well-known coaction on multiple polylogarithms. We illustrate the validity of our construction by explicitly determining the coaction on various types of hypergeometric ${}_{p+1}F_p$ and Appell functions. △ Less

Submitted 18 October, 2019; originally announced October 2019.

Generating Local Search Neighborhood with Synthesized Logic Programs

Authors: Mateusz Ślażyński, Salvador Abreu, Grzegorz J. Nalepa

Abstract: Local Search meta-heuristics have been proven a viable approach to solve difficult optimization problems. Their performance depends strongly on the search space landscape, as defined by a cost function and the selected neighborhood operators. In this paper we present a logic programming based framework, named Noodle, designed to generate bespoke Local Search neighborhoods tailored to specific dis… ▽ More Local Search meta-heuristics have been proven a viable approach to solve difficult optimization problems. Their performance depends strongly on the search space landscape, as defined by a cost function and the selected neighborhood operators. In this paper we present a logic programming based framework, named Noodle, designed to generate bespoke Local Search neighborhoods tailored to specific discrete optimization problems. The proposed system consists of a domain specific language, which is inspired by logic programming, as well as a genetic programming solver, based on the grammar evolution algorithm. We complement the description with a preliminary experimental evaluation, where we synthesize efficient neighborhood operators for the traveling salesman problem, some of which reproduce well-known results. △ Less

Submitted 18 September, 2019; originally announced September 2019.

Comments: In Proceedings ICLP 2019, arXiv:1909.07646

Journal ref: EPTCS 306, 2019, pp. 168-181

Pre-proceedings of the DECLARE 2019 Conference

Authors: Salvador Abreu, Petra Hofstedt, Ulrich John, Herbert Kuchen, Dietmar Seipel

Abstract: This volume constitutes the pre-proceedings of the DECLARE 2019 conference, held on September 9 to 13, 2019 at the University of Technology Cottbus - Senftenberg (Germany). Declarative programming is an advanced paradigm for the modeling and solving of complex problems. This method has attracted increased attention over the last decades, e.g., in the domains of data and knowledge engineering, da… ▽ More This volume constitutes the pre-proceedings of the DECLARE 2019 conference, held on September 9 to 13, 2019 at the University of Technology Cottbus - Senftenberg (Germany). Declarative programming is an advanced paradigm for the modeling and solving of complex problems. This method has attracted increased attention over the last decades, e.g., in the domains of data and knowledge engineering, databases, artificial intelligence, natural language processing, modeling and processing combinatorial problems, and for establishing systems for the web. The conference DECLARE 2019 aims at cross-fertilizing exchange of ideas and experiences among researches and students from the different communities interested in the foundations, implementation techniques, novel applications, and combinations of high-level, declarative programming and related areas. The technical program of the event included invited talks, presentations of refereed papers, and system demonstrations. DECLARE 2019 consisted of the sub-events INAP, WFLP, and WLP: INAP - 22nd International Conference on Applications of Declarative Programming and Knowledge Management WFLP - 27th International Workshop on Functional and (Constraint) Logic Programming WLP - 33rd Workshop on (Constraint) Logic Programming △ Less

Submitted 21 November, 2019; v1 submitted 11 September, 2019; originally announced September 2019.

Automated Architecture Design for Deep Neural Networks

Authors: Steven Abreu

Abstract: Machine learning has made tremendous progress in recent years and received large amounts of public attention. Though we are still far from designing a full artificially intelligent agent, machine learning has brought us many applications in which computers solve human learning tasks remarkably well. Much of this progress comes from a recent trend within machine learning, called deep learning. Deep… ▽ More Machine learning has made tremendous progress in recent years and received large amounts of public attention. Though we are still far from designing a full artificially intelligent agent, machine learning has brought us many applications in which computers solve human learning tasks remarkably well. Much of this progress comes from a recent trend within machine learning, called deep learning. Deep learning models are responsible for many state-of-the-art applications of machine learning. Despite their success, deep learning models are hard to train, very difficult to understand, and often times so complex that training is only possible on very large GPU clusters. Lots of work has been done on enabling neural networks to learn efficiently. However, the design and architecture of such neural networks is often done manually through trial and error and expert knowledge. This thesis inspects different approaches, existing and novel, to automate the design of deep feedforward neural networks in an attempt to create less complex models with good performance that take away the burden of deciding on an architecture and make it more efficient to design and train such deep networks. △ Less

Submitted 21 August, 2019; originally announced August 2019.

Comments: Undergraduate Thesis

Theory for the FCC-ee : Report on the 11th FCC-ee Workshop

Authors: A. Blondel, J. Gluza, S. Jadach, P. Janot, T. Riemann, S. Abreu, J. J. Aguilera-Verdugo, A. B. Arbuzov, J. Baglio, S. D. Bakshi, S. Banerjee, M. Beneke, C. Bobeth, C. Bogner, S. Bondarenko, S. Borowka, S. Braß, C. M. Carloni Calame, J. Chakrabortty, M. Chiesa, M. Chrzaszcz, D. d'Enterria, F. Domingo, J. Dormans, F. Driencourt-Mangin , et al. (61 additional authors not shown)

Abstract: The Future Circular Collider (FCC) at CERN, a proposed 100-km circular facility with several colliders in succession, culminates with a 100 TeV proton-proton collider. It offers a vast new domain of exploration in particle physics, with orders of magnitude advances in terms of Precision, Sensitivity and Energy. The implementation plan foresees, as a first step, an Electroweak Factory electron-posi… ▽ More The Future Circular Collider (FCC) at CERN, a proposed 100-km circular facility with several colliders in succession, culminates with a 100 TeV proton-proton collider. It offers a vast new domain of exploration in particle physics, with orders of magnitude advances in terms of Precision, Sensitivity and Energy. The implementation plan foresees, as a first step, an Electroweak Factory electron-positron collider. This high luminosity facility, operating between 90 and 365 GeV centre-of-mass energy, will study the heavy particles of the Standard Model, Z, W, Higgs, and top with unprecedented accuracy. The Electroweak Factory $e^+e^-$ collider constitutes a real challenge to the theory and to precision calculations, triggering the need for the development of new mathematical methods and software tools. A first workshop in 2018 had focused on the first FCC-ee stage, the Tera-Z, and confronted the theoretical status of precision Standard Model calculations on the Z-boson resonance to the experimental demands. The second workshop in January 2019, which is reported here, extended the scope to the next stages, with the production of W-bosons (FCC-ee-W), the Higgs boson (FCC-ee-H) and top quarks (FCC-ee-tt). In particular, the theoretical precision in the determination of the crucial input parameters, alpha_QED, alpha_QCD, M_W, m_t at the level of FCC-ee requirements is thoroughly discussed. The requirements on Standard Model theory calculations were spelled out, so as to meet the demanding accuracy of the FCC-ee experimental potential. The discussion of innovative methods and tools for multi-loop calculations was deepened. Furthermore, phenomenological analyses beyond the Standard Model were discussed, in particular the effective theory approaches. The reports of 2018 and 2019 serve as white papers of the workshop results and subsequent developments. △ Less

Submitted 19 May, 2020; v1 submitted 13 May, 2019; originally announced May 2019.

Comments: Published version

Journal ref: CERN Yellow Reports: Monographs, CERN-2020-003

Analytic Form of the Planar Two-Loop Five-Parton Scattering Amplitudes in QCD

Authors: S. Abreu, J. Dormans, F. Febres Cordero, H. Ita, B. Page, V. Sotnikov

Abstract: We present the analytic form of all leading-color two-loop five-parton helicity amplitudes in QCD. The results are analytically reconstructed from exact numerical evaluations over finite fields. Combining a judicious choice of variables with a new approach to the treatment of particle states in $D$ dimensions for the numerical evaluation of amplitudes, we obtain the analytic expressions with a mod… ▽ More We present the analytic form of all leading-color two-loop five-parton helicity amplitudes in QCD. The results are analytically reconstructed from exact numerical evaluations over finite fields. Combining a judicious choice of variables with a new approach to the treatment of particle states in $D$ dimensions for the numerical evaluation of amplitudes, we obtain the analytic expressions with a modest computational effort. Their systematic simplification using multivariate partial-fraction decomposition leads to a particularly compact form. Our results provide all two-loop amplitudes required for the calculation of next-to-next-to-leading order QCD corrections to the production of three jets at hadron colliders in the leading-color approximation. △ Less

Submitted 22 May, 2019; v1 submitted 1 April, 2019; originally announced April 2019.

Comments: v2: Minor text changes, matches published version

The two-loop five-point amplitude in $\mathcal N=8$ supergravity

Authors: Samuel Abreu, Lance J. Dixon, Enrico Herrmann, Ben Page, Mao Zeng

Abstract: We compute the symbol of the two-loop five-point amplitude in $\mathcal N=8$ supergravity. We write an ansatz for the amplitude whose rational prefactors are based on not only 4-dimensional leading singularities, but also $d$-dimensional ones, as the former are insufficient. Our novel $d$-dimensional unitarity-based approach to the systematic construction of an amplitude's rational structures is l… ▽ More We compute the symbol of the two-loop five-point amplitude in $\mathcal N=8$ supergravity. We write an ansatz for the amplitude whose rational prefactors are based on not only 4-dimensional leading singularities, but also $d$-dimensional ones, as the former are insufficient. Our novel $d$-dimensional unitarity-based approach to the systematic construction of an amplitude's rational structures is likely to have broader applications, for example to analogous QCD calculations. We fix parameters in the ansatz by performing numerical integration-by-parts reduction of the known integrand. We find that the two-loop five-point $\mathcal N=8$ supergravity amplitude is uniformly transcendental. We then verify the soft and collinear limits of the amplitude. There is considerable similarity with the corresponding amplitude for $\mathcal N=4$ super-Yang-Mills theory: all the rational prefactors are double copies of the Yang-Mills ones and the transcendental functions overlap to a large degree. As a byproduct, we find new relations between color-ordered loop amplitudes in $\mathcal N=4$ super-Yang-Mills theory. △ Less

Submitted 26 March, 2019; v1 submitted 24 January, 2019; originally announced January 2019.

Comments: 50 pages, 4 figures. v2: final journal version

Report number: CP3-19-06, IPhT-19/003, SLAC-PUB-17377, HU-EP-19/01

Journal ref: JHEP 03 (2019) 123

The two-loop five-point amplitude in $\mathcal{N} =4$ super-Yang-Mills theory

Authors: Samuel Abreu, Lance J. Dixon, Enrico Herrmann, Ben Page, Mao Zeng

Abstract: We compute the symbol of the two-loop five-point scattering amplitude in $\mathcal{N}$ = 4 super-Yang-Mills theory, including its full color dependence. This requires constructing the symbol of all two-loop five-point nonplanar massless master integrals, for which we give explicit results. We compute the symbol of the two-loop five-point scattering amplitude in $\mathcal{N}$ = 4 super-Yang-Mills theory, including its full color dependence. This requires constructing the symbol of all two-loop five-point nonplanar massless master integrals, for which we give explicit results. △ Less

Submitted 27 March, 2019; v1 submitted 20 December, 2018; originally announced December 2018.

Comments: 6pages, 1 figure, 1 table; v2: references added, fixed typos, flipped overall sign of amplitude in ancillary files; v3: references and typos fixed

Report number: CP3-18-31, IPhT-18/170, SLAC-PUB-17369

Journal ref: Phys. Rev. Lett. 122, 121603 (2019)

Analytic Form of the Planar Two-Loop Five-Gluon Scattering Amplitudes in QCD

Authors: S. Abreu, J. Dormans, F. Febres Cordero, H. Ita, B. Page

Abstract: We present the analytic form of the two-loop five-gluon scattering amplitudes in QCD for a complete set of independent helicity configurations of external gluons. These include the first analytic results for five-point two-loop amplitudes relevant for the computation of next-to-next-to-leading-order QCD corrections at hadron colliders. The results were obtained by reconstructing analytic expressio… ▽ More We present the analytic form of the two-loop five-gluon scattering amplitudes in QCD for a complete set of independent helicity configurations of external gluons. These include the first analytic results for five-point two-loop amplitudes relevant for the computation of next-to-next-to-leading-order QCD corrections at hadron colliders. The results were obtained by reconstructing analytic expressions from numerical evaluations. The complexity of the computation is reduced by exploiting physical and analytical properties of the amplitudes, employing a minimal basis of so-called pentagon functions that have recently been classified. △ Less

Submitted 11 December, 2018; originally announced December 2018.

Comments: Ancillary files with analytic expressions can be downloaded from https://www.qft.physik.uni-freiburg.de/ScatteringAmplitudes

Journal ref: Phys. Rev. Lett. 122, 082002 (2019)