-
Ideal-theoretic non-noetherianity of polynomial functors in positive characteristic
Authors:
Karthik Ganapathy
Abstract:
A long-standing open problem in representation stability is whether every finitely generated commutative algebra in the category of strict polynomial functors satisfies the noetherian property. In this paper, we resolve this problem negatively over fields of positive characteristic using ideas from invariant theory. Specifically, we consider the algebra $P$ of polarizations of elementary symmetric…
▽ More
A long-standing open problem in representation stability is whether every finitely generated commutative algebra in the category of strict polynomial functors satisfies the noetherian property. In this paper, we resolve this problem negatively over fields of positive characteristic using ideas from invariant theory. Specifically, we consider the algebra $P$ of polarizations of elementary symmetric polynomials inside the ring of all multisymmetric polynomials in $p \times \infty$ variables. We show $P$ is not noetherian based on two key facts: (1) the $p$-th power of every multisymmetric polynomial is in $P$ (our main technical result) and (2) the ring of multisymmetric polynomials is Frobenius split.
△ Less
Submitted 18 June, 2025;
originally announced June 2025.
-
Tensor invariants of finite and classical groups
Authors:
Karthik Ganapathy
Abstract:
Given a faithful representation $W$ of a finite group $G$, we classify when the invariant subring of the tensor algebra $T(W)$ is noetherian/finitely generated as a twisted commutative algebra (tca). For a representation $W$ of a classical group $G_{\mathbb{Z}}$, we show that the invariant subring $T(W_k)^{G_k}$ is a finitely generated tca for $k$ algebraically closed of sufficiently large charact…
▽ More
Given a faithful representation $W$ of a finite group $G$, we classify when the invariant subring of the tensor algebra $T(W)$ is noetherian/finitely generated as a twisted commutative algebra (tca). For a representation $W$ of a classical group $G_{\mathbb{Z}}$, we show that the invariant subring $T(W_k)^{G_k}$ is a finitely generated tca for $k$ algebraically closed of sufficiently large characteristics, assuming $W$ admits a good filtration over $\mathbb{Z}$. Finally, we introduce a categorical variant of the Gelfand--Kirillov dimension and compute its value to be $\binom{n+1}{2}$ for $T(\mathbb{C}^n)$ as a tca.
△ Less
Submitted 23 February, 2025;
originally announced February 2025.
-
Non-noetherian GL-algebras in characteristic two
Authors:
Karthik Ganapathy
Abstract:
Over fields of characteristic two, we construct an infinite ascending chain of GL-stable ideals in the coordinate ring of infinite skew-symmetric matrices. This construction provides the first known example of a non-noetherian GL-algebra, thereby resolving a long-standing open question in the area. Our results build on the work of Draisma, Krasilnikov, and Krone.
Over fields of characteristic two, we construct an infinite ascending chain of GL-stable ideals in the coordinate ring of infinite skew-symmetric matrices. This construction provides the first known example of a non-noetherian GL-algebra, thereby resolving a long-standing open question in the area. Our results build on the work of Draisma, Krasilnikov, and Krone.
△ Less
Submitted 14 August, 2024; v1 submitted 8 August, 2024;
originally announced August 2024.
-
Resolutions of symmetric ideals via stratifications of derived categories
Authors:
Karthik Ganapathy
Abstract:
We propose a method to unify various stability results about symmetric ideals in polynomial rings by stratifying related derived categories. We execute this idea for chains of $GL_n$-equivariant modules over an infinite field $k$ of positive characteristic. We prove the Le--Nagel--Nguyen--Römer conjectures for such sequences and obtain stability patterns in their resolutions as corollaries of our…
▽ More
We propose a method to unify various stability results about symmetric ideals in polynomial rings by stratifying related derived categories. We execute this idea for chains of $GL_n$-equivariant modules over an infinite field $k$ of positive characteristic. We prove the Le--Nagel--Nguyen--Römer conjectures for such sequences and obtain stability patterns in their resolutions as corollaries of our main result, which is a semiorthogonal decomposition for the bounded derived category of $GL_{\infty}$-equivariant modules over $S = k[x_1, x_2, \ldots, x_n, \ldots]$. Our method relies on finite generation results for certain local cohomology modules. We also outline approaches (i) to investigate Koszul duality for $S$-modules taking the Frobenius homomorphism (of $GL_{\infty}$) into account, and (ii) to recover and extend Murai's results about free resolutions of symmetric monomial ideals.
△ Less
Submitted 22 July, 2024;
originally announced July 2024.
-
GL-algebras in positive characteristic II: the polynomial ring
Authors:
Karthik Ganapathy
Abstract:
We study GL-equivariant modules over the infinite variable polynomial ring $S = k[x_1, x_2, ..., x_n, ...]$ with $k$ an infinite field of characteristic $p > 0$. We extend many of Sam--Snowden's far-reaching results from characteristic zero to this setting. For example, while the Castelnuovo--Mumford regularity of a finitely generated GL-equivariant $S$-module need not be finite in positive charac…
▽ More
We study GL-equivariant modules over the infinite variable polynomial ring $S = k[x_1, x_2, ..., x_n, ...]$ with $k$ an infinite field of characteristic $p > 0$. We extend many of Sam--Snowden's far-reaching results from characteristic zero to this setting. For example, while the Castelnuovo--Mumford regularity of a finitely generated GL-equivariant $S$-module need not be finite in positive characteristic, we show that the resolution still has finitely many "linear strands of higher slope".
The crux of this paper is two technical results. The first is an extension to positive characteristic of Snowden's recent linearization of Draisma's embedding theorem which we use to study the generic category of $S$-modules. The second is a Nagpal-type "shift theorem" about torsion $S$-modules for which we introduce certain categorifications of the Hasse derivative. These two results together allow us to obtain explicit generators for the derived category. In a follow-up paper, we also use these results to prove finiteness results for local cohomology modules.
△ Less
Submitted 18 July, 2024;
originally announced July 2024.
-
Self-Tuning Network Control Architectures with Joint Sensor and Actuator Selection
Authors:
Karthik Ganapathy,
Iman Shames,
Mathias Hudoba de Badyn,
Tyler Summers
Abstract:
We formulate a mathematical framework for designing a self-tuning network control architecture, and propose a computationally-feasible greedy algorithm for online architecture optimization. In this setting, the locations of active sensors and actuators in the network, as well as the feedback control policy are jointly adapted using all available information about the network states and dynamics to…
▽ More
We formulate a mathematical framework for designing a self-tuning network control architecture, and propose a computationally-feasible greedy algorithm for online architecture optimization. In this setting, the locations of active sensors and actuators in the network, as well as the feedback control policy are jointly adapted using all available information about the network states and dynamics to optimize a performance criterion. We show that the case with full-state feedback can be solved with dynamic programming, and in the linear-quadratic setting, the optimal cost functions and policies are piecewise quadratic and piecewise linear, respectively. Our framework is extended for joint sensor and actuator selection for dynamic output feedback control with both control performance and architecture costs. For large networks where exhaustive architecture search is prohibitive, we describe a greedy heuristic for actuator selection and propose a greedy swapping algorithm for joint sensor and actuator selection. Via numerical experiments, we demonstrate a dramatic performance improvement of greedy self-tuning architectures over fixed architectures. Our general formulation provides an extremely rich and challenging problem space with opportunities to apply a wide variety of approximation methods from stochastic control, system identification, reinforcement learning, and static architecture design for practical model-based control.
△ Less
Submitted 19 January, 2024;
originally announced February 2024.
-
Self-Tuning Network Control Architectures
Authors:
Tyler Summers,
Karthik Ganapathy,
Iman Shames,
Mathias Hudoba de Badyn
Abstract:
We formulate a general mathematical framework for self-tuning network control architecture design. This problem involves jointly adapting the locations of active sensors and actuators in the network and the feedback control policy to all available information about the time-varying network state and dynamics to optimize a performance criterion. We propose a general solution structure analogous to…
▽ More
We formulate a general mathematical framework for self-tuning network control architecture design. This problem involves jointly adapting the locations of active sensors and actuators in the network and the feedback control policy to all available information about the time-varying network state and dynamics to optimize a performance criterion. We propose a general solution structure analogous to the classical self-tuning regulator from adaptive control. We show that a special case with full-state feedback can be solved in principle with dynamic programming, and in the linear quadratic setting the optimal cost functions and policies are piecewise quadratic and piecewise linear, respectively. For large networks where exhaustive architecture search is prohibitive, we describe a greedy heuristic for joint architecture-policy design. We demonstrate in numerical experiments that self-tuning architectures can provide dramatically improved performance over fixed architectures. Our general formulation provides an extremely rich and challenging problem space with opportunities to apply a wide variety of approximation methods from stochastic control, system identification, reinforcement learning, and static architecture design.
△ Less
Submitted 16 January, 2023;
originally announced January 2023.
-
GL-algebras in positive characteristic I: the exterior algebra
Authors:
Karthik Ganapathy
Abstract:
We study the category of GL-equivariant modules over the infinite exterior algebra in positive characteristic. Our main structural result is a shift theorem a la Nagpal. Using this, we obtain a Church--Ellenberg type bound for the Castelnuovo--Mumford regularity. We also prove finiteness results for local cohomology.
We study the category of GL-equivariant modules over the infinite exterior algebra in positive characteristic. Our main structural result is a shift theorem a la Nagpal. Using this, we obtain a Church--Ellenberg type bound for the Castelnuovo--Mumford regularity. We also prove finiteness results for local cohomology.
△ Less
Submitted 5 July, 2024; v1 submitted 7 March, 2022;
originally announced March 2022.
-
History Data Driven Distributed Consensus in Networks
Authors:
Venkatraman Renganathan,
Angela Fontan,
Karthik Ganapathy
Abstract:
The association of weights in a distributed consensus protocol quantify the trust that an agent has on its neighbors in a network. An important problem in such networked systems is the uncertainty in the estimation of trust between neighboring agents, coupled with the losses arising from mistakenly associating wrong amounts of trust with different neighboring agents. We introduce a probabilistic a…
▽ More
The association of weights in a distributed consensus protocol quantify the trust that an agent has on its neighbors in a network. An important problem in such networked systems is the uncertainty in the estimation of trust between neighboring agents, coupled with the losses arising from mistakenly associating wrong amounts of trust with different neighboring agents. We introduce a probabilistic approach which uses the historical data collected in the network, to determine the level of trust between each agent. Specifically, using the finite history of the shared data between neighbors, we obtain a configuration which represents the confidence estimate of every neighboring agent's trustworthiness. Finally, we propose a History-Data-Driven (HDD) distributed consensus protocol which translates the computed configuration data into weights to be used in the consensus update. The approach using the historical data in the context of a distributed consensus setting marks the novel contribution of our paper.
△ Less
Submitted 18 February, 2022;
originally announced February 2022.
-
Analyzing the Machine Learning Conference Review Process
Authors:
David Tran,
Alex Valtchanov,
Keshav Ganapathy,
Raymond Feng,
Eric Slud,
Micah Goldblum,
Tom Goldstein
Abstract:
Mainstream machine learning conferences have seen a dramatic increase in the number of participants, along with a growing range of perspectives, in recent years. Members of the machine learning community are likely to overhear allegations ranging from randomness of acceptance decisions to institutional bias. In this work, we critically analyze the review process through a comprehensive study of pa…
▽ More
Mainstream machine learning conferences have seen a dramatic increase in the number of participants, along with a growing range of perspectives, in recent years. Members of the machine learning community are likely to overhear allegations ranging from randomness of acceptance decisions to institutional bias. In this work, we critically analyze the review process through a comprehensive study of papers submitted to ICLR between 2017 and 2020. We quantify reproducibility/randomness in review scores and acceptance decisions, and examine whether scores correlate with paper impact. Our findings suggest strong institutional bias in accept/reject decisions, even after controlling for paper quality. Furthermore, we find evidence for a gender gap, with female authors receiving lower scores, lower acceptance rates, and fewer citations per paper than their male counterparts. We conclude our work with recommendations for future conference organizers.
△ Less
Submitted 25 November, 2020; v1 submitted 24 November, 2020;
originally announced November 2020.
-
An Open Review of OpenReview: A Critical Analysis of the Machine Learning Conference Review Process
Authors:
David Tran,
Alex Valtchanov,
Keshav Ganapathy,
Raymond Feng,
Eric Slud,
Micah Goldblum,
Tom Goldstein
Abstract:
Mainstream machine learning conferences have seen a dramatic increase in the number of participants, along with a growing range of perspectives, in recent years. Members of the machine learning community are likely to overhear allegations ranging from randomness of acceptance decisions to institutional bias. In this work, we critically analyze the review process through a comprehensive study of pa…
▽ More
Mainstream machine learning conferences have seen a dramatic increase in the number of participants, along with a growing range of perspectives, in recent years. Members of the machine learning community are likely to overhear allegations ranging from randomness of acceptance decisions to institutional bias. In this work, we critically analyze the review process through a comprehensive study of papers submitted to ICLR between 2017 and 2020. We quantify reproducibility/randomness in review scores and acceptance decisions, and examine whether scores correlate with paper impact. Our findings suggest strong institutional bias in accept/reject decisions, even after controlling for paper quality. Furthermore, we find evidence for a gender gap, with female authors receiving lower scores, lower acceptance rates, and fewer citations per paper than their male counterparts. We conclude our work with recommendations for future conference organizers.
△ Less
Submitted 26 October, 2020; v1 submitted 10 October, 2020;
originally announced October 2020.
-
A Study of Genetic Algorithms for Hyperparameter Optimization of Neural Networks in Machine Translation
Authors:
Keshav Ganapathy
Abstract:
With neural networks having demonstrated their versatility and benefits, the need for their optimal performance is as prevalent as ever. A defining characteristic, hyperparameters, can greatly affect its performance. Thus engineers go through a process, tuning, to identify and implement optimal hyperparameters. That being said, excess amounts of manual effort are required for tuning network archit…
▽ More
With neural networks having demonstrated their versatility and benefits, the need for their optimal performance is as prevalent as ever. A defining characteristic, hyperparameters, can greatly affect its performance. Thus engineers go through a process, tuning, to identify and implement optimal hyperparameters. That being said, excess amounts of manual effort are required for tuning network architectures, training configurations, and preprocessing settings such as Byte Pair Encoding (BPE). In this study, we propose an automatic tuning method modeled after Darwin's Survival of the Fittest Theory via a Genetic Algorithm (GA). Research results show that the proposed method, a GA, outperforms a random selection of hyperparameters.
△ Less
Submitted 14 September, 2020;
originally announced September 2020.
-
Stillman's question for twisted commutative algebras
Authors:
Karthik Ganapathy
Abstract:
Let $\mathbf{A}_{n, m}$ be the polynomial ring $\text{Sym}(\mathbf{C}^n \otimes \mathbf{C}^m)$ with the natural action of $\mathbf{GL}_m(\mathbf{C})$. We construct a family of $\mathbf{GL}_m(\mathbf{C})$-stable ideals $J_{n, m}$ in $\mathbf{A}_{n, m}$, each equivariantly generated by one homogeneous polynomial of degree $2$. Using the Ananyan-Hochster principle, we show that the regularity of this…
▽ More
Let $\mathbf{A}_{n, m}$ be the polynomial ring $\text{Sym}(\mathbf{C}^n \otimes \mathbf{C}^m)$ with the natural action of $\mathbf{GL}_m(\mathbf{C})$. We construct a family of $\mathbf{GL}_m(\mathbf{C})$-stable ideals $J_{n, m}$ in $\mathbf{A}_{n, m}$, each equivariantly generated by one homogeneous polynomial of degree $2$. Using the Ananyan-Hochster principle, we show that the regularity of this family is unbounded. This negatively answers a question raised by Erman-Sam-Snowden on a generalization of Stillman's conjecture.
△ Less
Submitted 19 August, 2020; v1 submitted 6 July, 2020;
originally announced July 2020.
-
Extensible Component Based Architecture for FLASH, A Massively Parallel, Multiphysics Simulation Code
Authors:
A. Dubey,
L. B. Reid,
K. Weide,
K. Antypas,
M. K. Ganapathy,
K. Riley,
D. Sheeler,
A. Siegal
Abstract:
FLASH is a publicly available high performance application code which has evolved into a modular, extensible software system from a collection of unconnected legacy codes. FLASH has been successful because its capabilities have been driven by the needs of scientific applications, without compromising maintainability, performance, and usability. In its newest incarnation, FLASH3 consists of inter…
▽ More
FLASH is a publicly available high performance application code which has evolved into a modular, extensible software system from a collection of unconnected legacy codes. FLASH has been successful because its capabilities have been driven by the needs of scientific applications, without compromising maintainability, performance, and usability. In its newest incarnation, FLASH3 consists of inter-operable modules that can be combined to generate different applications. The FLASH architecture allows arbitrarily many alternative implementations of its components to co-exist and interchange with each other, resulting in greater flexibility. Further, a simple and elegant mechanism exists for customization of code functionality without the need to modify the core implementation of the source. A built-in unit test framework providing verifiability, combined with a rigorous software maintenance process, allow the code to operate simultaneously in the dual mode of production and development. In this paper we describe the FLASH3 architecture, with emphasis on solutions to the more challenging conflicts arising from solver complexity, portable performance requirements, and legacy codes. We also include results from user surveys conducted in 2005 and 2007, which highlight the success of the code.
△ Less
Submitted 24 July, 2009; v1 submitted 27 March, 2009;
originally announced March 2009.
-
A Tight Bound for the Lamplighter Problem
Authors:
Murali K. Ganapathy,
Prasad Tetali
Abstract:
We settle an open problem, raised by Y. Peres and D. Revelle, concerning the $L^2$ mixing time of the random walk on the lamplighter graph. We also provide general bounds relating the entropy decay of a Markov chain to the separation distance of the chain, and show that the lamplighter graphs once again provide examples of tightness of our results.
We settle an open problem, raised by Y. Peres and D. Revelle, concerning the $L^2$ mixing time of the random walk on the lamplighter graph. We also provide general bounds relating the entropy decay of a Markov chain to the separation distance of the chain, and show that the lamplighter graphs once again provide examples of tightness of our results.
△ Less
Submitted 10 October, 2006;
originally announced October 2006.
