-
Enhancing Item Tokenization for Generative Recommendation through Self-Improvement
Authors:
Runjin Chen,
Mingxuan Ju,
Ngoc Bui,
Dimosthenis Antypas,
Stanley Cai,
Xiaopeng Wu,
Leonardo Neves,
Zhangyang Wang,
Neil Shah,
Tong Zhao
Abstract:
Generative recommendation systems, driven by large language models (LLMs), present an innovative approach to predicting user preferences by modeling items as token sequences and generating recommendations in a generative manner. A critical challenge in this approach is the effective tokenization of items, ensuring that they are represented in a form compatible with LLMs. Current item tokenization…
▽ More
Generative recommendation systems, driven by large language models (LLMs), present an innovative approach to predicting user preferences by modeling items as token sequences and generating recommendations in a generative manner. A critical challenge in this approach is the effective tokenization of items, ensuring that they are represented in a form compatible with LLMs. Current item tokenization methods include using text descriptions, numerical strings, or sequences of discrete tokens. While text-based representations integrate seamlessly with LLM tokenization, they are often too lengthy, leading to inefficiencies and complicating accurate generation. Numerical strings, while concise, lack semantic depth and fail to capture meaningful item relationships. Tokenizing items as sequences of newly defined tokens has gained traction, but it often requires external models or algorithms for token assignment. These external processes may not align with the LLM's internal pretrained tokenization schema, leading to inconsistencies and reduced model performance. To address these limitations, we propose a self-improving item tokenization method that allows the LLM to refine its own item tokenizations during training process. Our approach starts with item tokenizations generated by any external model and periodically adjusts these tokenizations based on the LLM's learned patterns. Such alignment process ensures consistency between the tokenization and the LLM's internal understanding of the items, leading to more accurate recommendations. Furthermore, our method is simple to implement and can be integrated as a plug-and-play enhancement into existing generative recommendation systems. Experimental results on multiple datasets and using various initial tokenization strategies demonstrate the effectiveness of our method, with an average improvement of 8\% in recommendation performance.
△ Less
Submitted 22 December, 2024;
originally announced December 2024.
-
Automatic Extraction of Metaphoric Analogies from Literary Texts: Task Formulation, Dataset Construction, and Evaluation
Authors:
Joanne Boisson,
Zara Siddique,
Hsuvas Borkakoty,
Dimosthenis Antypas,
Luis Espinosa Anke,
Jose Camacho-Collados
Abstract:
Extracting metaphors and analogies from free text requires high-level reasoning abilities such as abstraction and language understanding. Our study focuses on the extraction of the concepts that form metaphoric analogies in literary texts. To this end, we construct a novel dataset in this domain with the help of domain experts. We compare the out-of-the-box ability of recent large language models…
▽ More
Extracting metaphors and analogies from free text requires high-level reasoning abilities such as abstraction and language understanding. Our study focuses on the extraction of the concepts that form metaphoric analogies in literary texts. To this end, we construct a novel dataset in this domain with the help of domain experts. We compare the out-of-the-box ability of recent large language models (LLMs) to structure metaphoric mappings from fragments of texts containing proportional analogies. The models are further evaluated on the generation of implicit elements of the analogy, which are indirectly suggested in the texts and inferred by human readers. The competitive results obtained by LLMs in our experiments are encouraging and open up new avenues such as automatically extracting analogies and metaphors from text instead of investing resources in domain experts to manually label data.
△ Less
Submitted 19 December, 2024;
originally announced December 2024.
-
Sensitive Content Classification in Social Media: A Holistic Resource and Evaluation
Authors:
Dimosthenis Antypas,
Indira Sen,
Carla Perez-Almendros,
Jose Camacho-Collados,
Francesco Barbieri
Abstract:
The detection of sensitive content in large datasets is crucial for ensuring that shared and analysed data is free from harmful material. However, current moderation tools, such as external APIs, suffer from limitations in customisation, accuracy across diverse sensitive categories, and privacy concerns. Additionally, existing datasets and open-source models focus predominantly on toxic language,…
▽ More
The detection of sensitive content in large datasets is crucial for ensuring that shared and analysed data is free from harmful material. However, current moderation tools, such as external APIs, suffer from limitations in customisation, accuracy across diverse sensitive categories, and privacy concerns. Additionally, existing datasets and open-source models focus predominantly on toxic language, leaving gaps in detecting other sensitive categories such as substance abuse or self-harm. In this paper, we put forward a unified dataset tailored for social media content moderation across six sensitive categories: conflictual language, profanity, sexually explicit material, drug-related content, self-harm, and spam. By collecting and annotating data with consistent retrieval strategies and guidelines, we address the shortcomings of previous focalised research. Our analysis demonstrates that fine-tuning large language models (LLMs) on this novel dataset yields significant improvements in detection performance compared to open off-the-shelf models such as LLaMA, and even proprietary OpenAI models, which underperform by 10-15% overall. This limitation is even more pronounced on popular moderation APIs, which cannot be easily tailored to specific sensitive content categories, among others.
△ Less
Submitted 6 December, 2024; v1 submitted 29 November, 2024;
originally announced November 2024.
-
Studies of parity violation in atoms
Authors:
Stefanos Nanos,
Iraklis Papigkiotis,
Dionysios Antypas
Abstract:
Studies of the effects of the weak interaction in atomic systems provide tests of the Standard Model of particle physics, and explore physics scenarios beyond the Standard Model. In addition, these studies can offer valuable insights into low-energy nuclear physics. We provide an overview of the field of atomic parity violation, and discuss implications to nuclear and particle physics, and ongoing…
▽ More
Studies of the effects of the weak interaction in atomic systems provide tests of the Standard Model of particle physics, and explore physics scenarios beyond the Standard Model. In addition, these studies can offer valuable insights into low-energy nuclear physics. We provide an overview of the field of atomic parity violation, and discuss implications to nuclear and particle physics, and ongoing experimental efforts. Furthermore, we present our plans for precision measurements of the signatures of the weak interaction in atomic ytterbium.
△ Less
Submitted 2 November, 2024;
originally announced November 2024.
-
Multilingual Topic Classification in X: Dataset and Analysis
Authors:
Dimosthenis Antypas,
Asahi Ushio,
Francesco Barbieri,
Jose Camacho-Collados
Abstract:
In the dynamic realm of social media, diverse topics are discussed daily, transcending linguistic boundaries. However, the complexities of understanding and categorising this content across various languages remain an important challenge with traditional techniques like topic modelling often struggling to accommodate this multilingual diversity. In this paper, we introduce X-Topic, a multilingual…
▽ More
In the dynamic realm of social media, diverse topics are discussed daily, transcending linguistic boundaries. However, the complexities of understanding and categorising this content across various languages remain an important challenge with traditional techniques like topic modelling often struggling to accommodate this multilingual diversity. In this paper, we introduce X-Topic, a multilingual dataset featuring content in four distinct languages (English, Spanish, Japanese, and Greek), crafted for the purpose of tweet topic classification. Our dataset includes a wide range of topics, tailored for social media content, making it a valuable resource for scientists and professionals working on cross-linguistic analysis, the development of robust multilingual models, and computational scientists studying online dialogue. Finally, we leverage X-Topic to perform a comprehensive cross-linguistic and multilingual analysis, and compare the capabilities of current general- and domain-specific language models.
△ Less
Submitted 3 October, 2024;
originally announced October 2024.
-
Coherent amplitude modulation of continuous-wave light in cesium vapor
Authors:
X. Zhang,
J. B. Kim,
D. Antypas
Abstract:
We report on observations of coherent, sustained oscillations in the absorption of continuous-wave light at 388 nm that excites the $6S_{1/2}\rightarrow 8P_{3/2}$ transition in cesium vapor. The oscillation frequency is close to the spacing of hyperfine levels of the $8P_{3/2}$ level that are excited simultaneously by the 388 nm field. We observe threshold behavior of the oscillation amplitude wit…
▽ More
We report on observations of coherent, sustained oscillations in the absorption of continuous-wave light at 388 nm that excites the $6S_{1/2}\rightarrow 8P_{3/2}$ transition in cesium vapor. The oscillation frequency is close to the spacing of hyperfine levels of the $8P_{3/2}$ level that are excited simultaneously by the 388 nm field. We observe threshold behavior of the oscillation amplitude with pump power, and suggest that the effect is associated with infrared directional emission due to amplified spontaneous emission from the $8P_{3/2}\rightarrow 8S_{1/2}$ transition, that is assisted by retro-reflections from the cell windows. The effect may be used to probe a lasing process in an atomic vapor, by checking the temporal properties of the pump field transmitted through the vapor.
△ Less
Submitted 15 June, 2024;
originally announced June 2024.
-
BLEnD: A Benchmark for LLMs on Everyday Knowledge in Diverse Cultures and Languages
Authors:
Junho Myung,
Nayeon Lee,
Yi Zhou,
Jiho Jin,
Rifki Afina Putri,
Dimosthenis Antypas,
Hsuvas Borkakoty,
Eunsu Kim,
Carla Perez-Almendros,
Abinew Ali Ayele,
Víctor Gutiérrez-Basulto,
Yazmín Ibáñez-García,
Hwaran Lee,
Shamsuddeen Hassan Muhammad,
Kiwoong Park,
Anar Sabuhi Rzayev,
Nina White,
Seid Muhie Yimam,
Mohammad Taher Pilehvar,
Nedjma Ousidhoum,
Jose Camacho-Collados,
Alice Oh
Abstract:
Large language models (LLMs) often lack culture-specific knowledge of daily life, especially across diverse regions and non-English languages. Existing benchmarks for evaluating LLMs' cultural sensitivities are limited to a single language or collected from online sources such as Wikipedia, which do not reflect the mundane everyday lifestyles of diverse regions. That is, information about the food…
▽ More
Large language models (LLMs) often lack culture-specific knowledge of daily life, especially across diverse regions and non-English languages. Existing benchmarks for evaluating LLMs' cultural sensitivities are limited to a single language or collected from online sources such as Wikipedia, which do not reflect the mundane everyday lifestyles of diverse regions. That is, information about the food people eat for their birthday celebrations, spices they typically use, musical instruments youngsters play, or the sports they practice in school is common cultural knowledge but uncommon in easily collected online sources, especially for underrepresented cultures. To address this issue, we introduce BLEnD, a hand-crafted benchmark designed to evaluate LLMs' everyday knowledge across diverse cultures and languages. BLEnD comprises 52.6k question-answer pairs from 16 countries/regions, in 13 different languages, including low-resource ones such as Amharic, Assamese, Azerbaijani, Hausa, and Sundanese. We construct the benchmark to include two formats of questions: short-answer and multiple-choice. We show that LLMs perform better for cultures that are highly represented online, with a maximum 57.34% difference in GPT-4, the best-performing model, in the short-answer format. For cultures represented by mid-to-high-resource languages, LLMs perform better in their local languages, but for cultures represented by low-resource languages, LLMs perform better in English than the local languages. We make our dataset publicly available at: https://github.com/nlee0212/BLEnD.
△ Less
Submitted 15 January, 2025; v1 submitted 14 June, 2024;
originally announced June 2024.
-
Words as Trigger Points in Social Media Discussions
Authors:
Dimosthenis Antypas,
Christian Arnold,
Jose Camacho-Collados,
Nedjma Ousidhoum,
Carla Perez Almendros
Abstract:
Trigger points, introduced by Mau et al . [30], are rooted in theories of affective political identity and relate to deeply lying beliefs about moral expectations and social dispositions. Examining trigger points in online discussions helps understand why and when social media users engage in disagreements or affective political deliberations. This opens the door to modelling social media user eng…
▽ More
Trigger points, introduced by Mau et al . [30], are rooted in theories of affective political identity and relate to deeply lying beliefs about moral expectations and social dispositions. Examining trigger points in online discussions helps understand why and when social media users engage in disagreements or affective political deliberations. This opens the door to modelling social media user engagement more effectively and studying the conditions and causal mechanisms that lead to adverse reactions, hate speech, and abusive language in online debates.
△ Less
Submitted 15 October, 2024; v1 submitted 16 May, 2024;
originally announced May 2024.
-
SuperTweetEval: A Challenging, Unified and Heterogeneous Benchmark for Social Media NLP Research
Authors:
Dimosthenis Antypas,
Asahi Ushio,
Francesco Barbieri,
Leonardo Neves,
Kiamehr Rezaee,
Luis Espinosa-Anke,
Jiaxin Pei,
Jose Camacho-Collados
Abstract:
Despite its relevance, the maturity of NLP for social media pales in comparison with general-purpose models, metrics and benchmarks. This fragmented landscape makes it hard for the community to know, for instance, given a task, which is the best performing model and how it compares with others. To alleviate this issue, we introduce a unified benchmark for NLP evaluation in social media, SuperTweet…
▽ More
Despite its relevance, the maturity of NLP for social media pales in comparison with general-purpose models, metrics and benchmarks. This fragmented landscape makes it hard for the community to know, for instance, given a task, which is the best performing model and how it compares with others. To alleviate this issue, we introduce a unified benchmark for NLP evaluation in social media, SuperTweetEval, which includes a heterogeneous set of tasks and datasets combined, adapted and constructed from scratch. We benchmarked the performance of a wide range of models on SuperTweetEval and our results suggest that, despite the recent advances in language modelling, social media remains challenging.
△ Less
Submitted 23 October, 2023;
originally announced October 2023.
-
Robust Hate Speech Detection in Social Media: A Cross-Dataset Empirical Evaluation
Authors:
Dimosthenis Antypas,
Jose Camacho-Collados
Abstract:
The automatic detection of hate speech online is an active research area in NLP. Most of the studies to date are based on social media datasets that contribute to the creation of hate speech detection models trained on them. However, data creation processes contain their own biases, and models inherently learn from these dataset-specific biases. In this paper, we perform a large-scale cross-datase…
▽ More
The automatic detection of hate speech online is an active research area in NLP. Most of the studies to date are based on social media datasets that contribute to the creation of hate speech detection models trained on them. However, data creation processes contain their own biases, and models inherently learn from these dataset-specific biases. In this paper, we perform a large-scale cross-dataset comparison where we fine-tune language models on different hate speech detection datasets. This analysis shows how some datasets are more generalisable than others when used as training data. Crucially, our experiments show how combining hate speech detection datasets can contribute to the development of robust hate speech detection models. This robustness holds even when controlling by data size and compared with the best individual datasets.
△ Less
Submitted 4 July, 2023;
originally announced July 2023.
-
Search for ultralight dark matter with spectroscopy of radio-frequency atomic transitions
Authors:
Xue Zhang,
Abhishek Banerjee,
Mahapan Leyser,
Gilad Perez,
Stephan Schiller,
Dmitry Budker,
Dionysios Antypas
Abstract:
The effects of scalar and pseudoscalar ultralight bosonic dark matter (UBDM) were searched for by comparing the frequency of a quartz oscillator to that of a hyperfine-structure transition in $^{87}$Rb, and an electronic transition in $^{164}$Dy. We constrain linear interactions between a scalar UBDM field and Standard-Model (SM) fields for an underlying UBDM particle mass in the range…
▽ More
The effects of scalar and pseudoscalar ultralight bosonic dark matter (UBDM) were searched for by comparing the frequency of a quartz oscillator to that of a hyperfine-structure transition in $^{87}$Rb, and an electronic transition in $^{164}$Dy. We constrain linear interactions between a scalar UBDM field and Standard-Model (SM) fields for an underlying UBDM particle mass in the range $1\times10^{-17}-8.3\times10^{-13} $ eV and quadratic interactions between a pseudoscalar UBDM field and SM fields in the range $5\times10^{-18}- 4.1\times10^{-13} $ eV. Within regions of the respective ranges, our constraints on linear interactions significantly improve on results from previous, direct searches for oscillations in atomic parameters, while constraints on quadratic interactions surpass limits imposed by such direct searches as well as by astrophysical observations.
△ Less
Submitted 8 December, 2022;
originally announced December 2022.
-
Twitter Topic Classification
Authors:
Dimosthenis Antypas,
Asahi Ushio,
Jose Camacho-Collados,
Leonardo Neves,
Vítor Silva,
Francesco Barbieri
Abstract:
Social media platforms host discussions about a wide variety of topics that arise everyday. Making sense of all the content and organising it into categories is an arduous task. A common way to deal with this issue is relying on topic modeling, but topics discovered using this technique are difficult to interpret and can differ from corpus to corpus. In this paper, we present a new task based on t…
▽ More
Social media platforms host discussions about a wide variety of topics that arise everyday. Making sense of all the content and organising it into categories is an arduous task. A common way to deal with this issue is relying on topic modeling, but topics discovered using this technique are difficult to interpret and can differ from corpus to corpus. In this paper, we present a new task based on tweet topic classification and release two associated datasets. Given a wide range of topics covering the most important discussion points in social media, we provide training and testing data from recent time periods that can be used to evaluate tweet classification models. Moreover, we perform a quantitative evaluation and analysis of current general- and domain-specific language models on the task, which provide more insights on the challenges and nature of the task.
△ Less
Submitted 20 September, 2022;
originally announced September 2022.
-
TweetNLP: Cutting-Edge Natural Language Processing for Social Media
Authors:
Jose Camacho-Collados,
Kiamehr Rezaee,
Talayeh Riahi,
Asahi Ushio,
Daniel Loureiro,
Dimosthenis Antypas,
Joanne Boisson,
Luis Espinosa-Anke,
Fangyu Liu,
Eugenio Martínez-Cámara,
Gonzalo Medina,
Thomas Buhrmann,
Leonardo Neves,
Francesco Barbieri
Abstract:
In this paper we present TweetNLP, an integrated platform for Natural Language Processing (NLP) in social media. TweetNLP supports a diverse set of NLP tasks, including generic focus areas such as sentiment analysis and named entity recognition, as well as social media-specific tasks such as emoji prediction and offensive language identification. Task-specific systems are powered by reasonably-siz…
▽ More
In this paper we present TweetNLP, an integrated platform for Natural Language Processing (NLP) in social media. TweetNLP supports a diverse set of NLP tasks, including generic focus areas such as sentiment analysis and named entity recognition, as well as social media-specific tasks such as emoji prediction and offensive language identification. Task-specific systems are powered by reasonably-sized Transformer-based language models specialized on social media text (in particular, Twitter) which can be run without the need for dedicated hardware or cloud services. The main contributions of TweetNLP are: (1) an integrated Python library for a modern toolkit supporting social media analysis using our various task-specific models adapted to the social domain; (2) an interactive online demo for codeless experimentation using our models; and (3) a tutorial covering a wide variety of typical social media applications.
△ Less
Submitted 25 October, 2022; v1 submitted 29 June, 2022;
originally announced June 2022.
-
New Horizons: Scalar and Vector Ultralight Dark Matter
Authors:
D. Antypas,
A. Banerjee,
C. Bartram,
M. Baryakhtar,
J. Betz,
J. J. Bollinger,
C. Boutan,
D. Bowring,
D. Budker,
D. Carney,
G. Carosi,
S. Chaudhuri,
S. Cheong,
A. Chou,
M. D. Chowdhury,
R. T. Co,
J. R. Crespo López-Urrutia,
M. Demarteau,
N. DePorzio,
A. V. Derbin,
T. Deshpande,
M. D. Chowdhury,
L. Di Luzio,
A. Diaz-Morcillo,
J. M. Doyle
, et al. (104 additional authors not shown)
Abstract:
The last decade has seen unprecedented effort in dark matter model building at all mass scales coupled with the design of numerous new detection strategies. Transformative advances in quantum technologies have led to a plethora of new high-precision quantum sensors and dark matter detection strategies for ultralight ($<10\,$eV) bosonic dark matter that can be described by an oscillating classical,…
▽ More
The last decade has seen unprecedented effort in dark matter model building at all mass scales coupled with the design of numerous new detection strategies. Transformative advances in quantum technologies have led to a plethora of new high-precision quantum sensors and dark matter detection strategies for ultralight ($<10\,$eV) bosonic dark matter that can be described by an oscillating classical, largely coherent field. This white paper focuses on searches for wavelike scalar and vector dark matter candidates.
△ Less
Submitted 28 March, 2022;
originally announced March 2022.
-
Negativity Spreads Faster: A Large-Scale Multilingual Twitter Analysis on the Role of Sentiment in Political Communication
Authors:
Dimosthenis Antypas,
Alun Preece,
Jose Camacho-Collados
Abstract:
Social media has become extremely influential when it comes to policy making in modern societies, especially in the western world, where platforms such as Twitter allow users to follow politicians, thus making citizens more involved in political discussion. In the same vein, politicians use Twitter to express their opinions, debate among others on current topics and promote their political agendas…
▽ More
Social media has become extremely influential when it comes to policy making in modern societies, especially in the western world, where platforms such as Twitter allow users to follow politicians, thus making citizens more involved in political discussion. In the same vein, politicians use Twitter to express their opinions, debate among others on current topics and promote their political agendas aiming to influence voter behaviour. In this paper, we attempt to analyse tweets of politicians from three European countries and explore the virality of their tweets. Previous studies have shown that tweets conveying negative sentiment are likely to be retweeted more frequently. By utilising state-of-the-art pre-trained language models, we performed sentiment analysis on hundreds of thousands of tweets collected from members of parliament in Greece, Spain and the United Kingdom, including devolved administrations. We achieved this by systematically exploring and analysing the differences between influential and less popular tweets. Our analysis indicates that politicians' negatively charged tweets spread more widely, especially in more recent times, and highlights interesting differences between political parties as well as between politicians and the general population.
△ Less
Submitted 3 April, 2023; v1 submitted 1 February, 2022;
originally announced February 2022.
-
Improved bounds on ultralight scalar dark matter in the radio-frequency range
Authors:
Oleg Tretiak,
Xue Zhang,
Nataniel L. Figueroa,
Dionysios Antypas,
Andrea Brogna,
Abhishek Banerjee,
Gilad Perez,
Dmitry Budker
Abstract:
We present a search for fundamental constant oscillations in the range $20$~kHz-$100$ MHz, that may arise within models for ultralight dark matter (UDM). Using two independent, significantly upgraded optical-spectroscopy apparatus, we achieve up to $\times$1000 greater sensitivity in the search relative to previous work. We report no observation of UDM and thus constrain respective couplings to el…
▽ More
We present a search for fundamental constant oscillations in the range $20$~kHz-$100$ MHz, that may arise within models for ultralight dark matter (UDM). Using two independent, significantly upgraded optical-spectroscopy apparatus, we achieve up to $\times$1000 greater sensitivity in the search relative to previous work. We report no observation of UDM and thus constrain respective couplings to electrons and photons within the investigated UDM particle mass range $8\cdot 10^{-11}-4\cdot 10^{-7}$ eV. The constraints significantly exceed previously set bounds, and as we show, may surpass in future experiments those provided by equivalence-principle experiments in a specific case regarding the combination of UDM couplings probed by the latter.
△ Less
Submitted 13 January, 2022; v1 submitted 6 January, 2022;
originally announced January 2022.
-
Search for oscillations of fundamental constants using molecular spectroscopy
Authors:
R. Oswald,
A. Nevsky,
V. Vogt,
S. Schiller,
N. L. Figueroa,
K. Zhang,
O. Tretiak,
D. Antypas,
D. Budker,
A. Banerjee,
G. Perez
Abstract:
A possible implication of an ultralight dark matter (UDM) field interacting wibeginth the Standard Model (SM) degrees of freedom is oscillations of fundamental constants. Here, we establish direct experimental bounds on the coupling of an oscillating UDM field to the up, down, and strange quarks and to the gluons, for oscillation frequencies between 10 Hz and 10^8 Hz. We employ spectroscopic exper…
▽ More
A possible implication of an ultralight dark matter (UDM) field interacting wibeginth the Standard Model (SM) degrees of freedom is oscillations of fundamental constants. Here, we establish direct experimental bounds on the coupling of an oscillating UDM field to the up, down, and strange quarks and to the gluons, for oscillation frequencies between 10 Hz and 10^8 Hz. We employ spectroscopic experiments that take advantage of the dependence of molecular transition frequencies on the nuclear masses. Our results apply to previously unexplored frequency bands, and improve on existing bounds at frequencies > 5 MHz. We identify a sector of UDM - SM coupling space where the bounds from Equivalence Principle tests may be challenged by next-generation experiments of the present kind.
△ Less
Submitted 12 November, 2021;
originally announced November 2021.
-
Precision determination of isotope shifts in ytterbium and implications for new physics
Authors:
N. L. Figueroa,
J. C. Berengut,
V. A. Dzuba,
V. V. Flambaum,
D. Budker,
D. Antypas
Abstract:
We report measurements of isotope shifts for the five spinless Yb isotopes on the $6s^2\,^1\textrm{S}_0 \rightarrow 5d6s\,^1\textrm{D}_2$ transition using Doppler-free two-photon spectroscopy. We combine these data with existing measurements on two transitions in Yb$^+$ [Phys. Rev. Lett. 125, 123002 (2020)], where deviation from King-plot linearity showed hints of a new bosonic force carrier at th…
▽ More
We report measurements of isotope shifts for the five spinless Yb isotopes on the $6s^2\,^1\textrm{S}_0 \rightarrow 5d6s\,^1\textrm{D}_2$ transition using Doppler-free two-photon spectroscopy. We combine these data with existing measurements on two transitions in Yb$^+$ [Phys. Rev. Lett. 125, 123002 (2020)], where deviation from King-plot linearity showed hints of a new bosonic force carrier at the 3$σ$ level. The combined data strongly reduces the significance of the new-physics signal. We show that the observed nonlinearity in the joint Yb/Yb$^+$ King-plot analysis can be accounted for by the deformation of the Yb nuclei.
△ Less
Submitted 31 January, 2022; v1 submitted 2 November, 2021;
originally announced November 2021.
-
Deriving Disinformation Insights from Geolocalized Twitter Callouts
Authors:
David Tuxworth,
Dimosthenis Antypas,
Luis Espinosa-Anke,
Jose Camacho-Collados,
Alun Preece,
David Rogers
Abstract:
This paper demonstrates a two-stage method for deriving insights from social media data relating to disinformation by applying a combination of geospatial classification and embedding-based language modelling across multiple languages. In particular, the analysis in centered on Twitter and disinformation for three European languages: English, French and Spanish. Firstly, Twitter data is classified…
▽ More
This paper demonstrates a two-stage method for deriving insights from social media data relating to disinformation by applying a combination of geospatial classification and embedding-based language modelling across multiple languages. In particular, the analysis in centered on Twitter and disinformation for three European languages: English, French and Spanish. Firstly, Twitter data is classified into European and non-European sets using BERT. Secondly, Word2vec is applied to the classified texts resulting in Eurocentric, non-Eurocentric and global representations of the data for the three target languages. This comparative analysis demonstrates not only the efficacy of the classification method but also highlights geographic, temporal and linguistic differences in the disinformation-related media. Thus, the contributions of the work are threefold: (i) a novel language-independent transformer-based geolocation method; (ii) an analytical approach that exploits lexical specificity and word embeddings to interrogate user-generated content; and (iii) a dataset of 36 million disinformation related tweets in English, French and Spanish.
△ Less
Submitted 6 August, 2021;
originally announced August 2021.
-
Probing fast oscillating scalar dark matter with atoms and molecules
Authors:
Dionysios Antypas,
Oleg Tretiak,
Ke Zhang,
Antoine Garcon,
Gilad Perez,
Mikhail G. Kozlov,
Stephan Schiller,
Dmitry Budker
Abstract:
Light scalar Dark Matter with scalar couplings to matter is expected within several scenarios to induce variations in the fundamental constants of nature. Such variations can be searched for, among other ways, via atomic spectroscopy. Sensitive atomic observables arise primarily due to possible changes in the fine-structure constant or the electron mass. Most of the searches to date have focused o…
▽ More
Light scalar Dark Matter with scalar couplings to matter is expected within several scenarios to induce variations in the fundamental constants of nature. Such variations can be searched for, among other ways, via atomic spectroscopy. Sensitive atomic observables arise primarily due to possible changes in the fine-structure constant or the electron mass. Most of the searches to date have focused on slow variations of the constants (i.e. modulation frequencies $<$ 1 Hz). In a recent experiment \mbox{[Phys. Rev. Lett. 123, 141102 (2019)]} called WReSL (Weekend Relaxion-Search Laboratory), we reported on a direct search for rapid variations in the radio-frequency band. Such a search is particularly motivated within a class of relaxion Dark Matter models. We discuss the WReSL experiment, report on progress towards improved measurements of rapid fundamental constant variations, and discuss the planned extension of the work to molecules, in which rapid variations of the nuclear mass can be sensitively searched for.
△ Less
Submitted 28 January, 2021; v1 submitted 2 December, 2020;
originally announced December 2020.
-
Fast apparent oscillations of fundamental constants
Authors:
Dionysios Antypas,
Dmitry Budker,
Victor V. Flambaum,
Mikhail G. Kozlov,
Gilad Perez,
Jun Ye
Abstract:
Precision spectroscopy of atoms and molecules allows one to search for and to put stringent limits on the variation of fundamental constants. These experiments are typically interpreted in terms of variations of the fine structure constant $α$ and the electron to proton mass ratio $μ=m_e/m_p$. Atomic spectroscopy is usually less sensitive to other fundamental constants, unless the hyperfine struct…
▽ More
Precision spectroscopy of atoms and molecules allows one to search for and to put stringent limits on the variation of fundamental constants. These experiments are typically interpreted in terms of variations of the fine structure constant $α$ and the electron to proton mass ratio $μ=m_e/m_p$. Atomic spectroscopy is usually less sensitive to other fundamental constants, unless the hyperfine structure of atomic levels is studied. However, the number of possible dimensionless constants increases when we allow for fast variations of the constants, where "fast" is determined by the time scale of the response of the studied species or experimental apparatus used. In this case, the relevant dimensionless quantity is, for example, the ratio $m_e/\langle m_e \rangle$ and $\langle m_e \rangle$ is the time average. In this sense, one may say that the experimental signal depends on the variation of dimensionful constants ($m_e$ in this example).
△ Less
Submitted 14 February, 2020; v1 submitted 3 December, 2019;
originally announced December 2019.
-
Polychromatic, continuous-wave mirrorless lasing from monochromatic pumping of cesium vapor
Authors:
D. Antypas,
O. Tretiak,
D. Budker,
A. Akulshin
Abstract:
We report on studies of simultaneous continuous-wave mirrorless lasing on multiple optical transitions, realized by pumping hot cesium vapor with laser light resonant with the 6$S_{1/2}\rightarrow 8$P$_{3/2}$ transition. The multiplicity of decay paths for the excited atoms to their ground state is responsible for the emergence of lasing in a number of transitions, observed here in at least seven…
▽ More
We report on studies of simultaneous continuous-wave mirrorless lasing on multiple optical transitions, realized by pumping hot cesium vapor with laser light resonant with the 6$S_{1/2}\rightarrow 8$P$_{3/2}$ transition. The multiplicity of decay paths for the excited atoms to their ground state is responsible for the emergence of lasing in a number of transitions, observed here in at least seven wavelengths in the infrared (IR), and at two wavelengths in the blue. We study the properties of the fields generated in the cesium vapor, such as optical power, directionality and optical linewidth.
△ Less
Submitted 26 June, 2019; v1 submitted 16 May, 2019;
originally announced May 2019.
-
Scalar dark matter in the radio-frequency band: atomic-spectroscopy search results
Authors:
D. Antypas,
O. Tretiak,
A. Garcon,
R. Ozeri,
G. Perez,
D. Budker
Abstract:
Among the prominent candidates for dark matter are bosonic fields with small scalar couplings to the Standard-Model particles. Several techniques are employed to search for such couplings and the current best constraints are derived from tests of gravity or atomic probes. In experiments employing atoms, observables would arise from expected dark-matter-induced oscillations in the fundamental const…
▽ More
Among the prominent candidates for dark matter are bosonic fields with small scalar couplings to the Standard-Model particles. Several techniques are employed to search for such couplings and the current best constraints are derived from tests of gravity or atomic probes. In experiments employing atoms, observables would arise from expected dark-matter-induced oscillations in the fundamental constants of nature. These studies are primarily sensitive to underlying particle masses below $10^{-14}$ eV. We present a method to search for fast oscillations of fundamental constants using atomic spectroscopy in cesium vapor. We demonstrate sensitivity to scalar interactions of dark matter associated with a particle mass in the range $8\cdot10^{-11}$ to $4\cdot 10^{-7}$ eV. In this range our experiment yields constraints on such interactions, which within the framework of an astronomical-size dark matter structure, are comparable with, or better than, those provided by experiments probing deviations from the law of gravity.
△ Less
Submitted 4 August, 2019; v1 submitted 8 May, 2019;
originally announced May 2019.
-
Isotopic variation of parity violation in atomic ytterbium: method of measurements and analysis of systematic effects
Authors:
D. Antypas,
A. M. Fabricant,
J. E. Stalnaker,
K. Tsigutkin,
V. V. Flambaum,
D. Budker
Abstract:
We present a detailed description of experimental studies of the parity violation effect in an isotopic chain of atomic ytterbium (Yb), whose results were reported in a recent Letter [Antypas et al., Nat. Phys. 15, 120 (2019)]. We discuss the principle of these measurements, made on the Yb 6s$^2$ $^1$S$_0 \rightarrow $5d6s $^3$D$_1$ optical transition at 408 nm, describe the experimental apparatus…
▽ More
We present a detailed description of experimental studies of the parity violation effect in an isotopic chain of atomic ytterbium (Yb), whose results were reported in a recent Letter [Antypas et al., Nat. Phys. 15, 120 (2019)]. We discuss the principle of these measurements, made on the Yb 6s$^2$ $^1$S$_0 \rightarrow $5d6s $^3$D$_1$ optical transition at 408 nm, describe the experimental apparatus, and give a detailed account of our studies of systematic effects in the experiment. Our results offer the first direct observation of the isotopic variation in the atomic parity violation effect, a variation which is in agreement with the prediction of the Standard Model. These measurements are used to constrain electron-proton and electron-neutron interactions, mediated by a light $Z'$ boson.
△ Less
Submitted 16 May, 2019; v1 submitted 14 March, 2019;
originally announced March 2019.
-
Dependence of atomic parity-violation effects on neutron skins and new physics
Authors:
A. V. Viatkina,
D. Antypas,
M. G. Kozlov,
D. Budker,
V. V. Flambaum
Abstract:
We estimate the relative contribution of nuclear structure and new physics couplings to the parity non-conserving spin-independent effects in atomic systems, for both single isotopes and isotopic ratios. General expressions are presented to assess the sensitivity of isotopic ratios to neutron skins and to couplings beyond standard model at tree level. The specific coefficients for these contributi…
▽ More
We estimate the relative contribution of nuclear structure and new physics couplings to the parity non-conserving spin-independent effects in atomic systems, for both single isotopes and isotopic ratios. General expressions are presented to assess the sensitivity of isotopic ratios to neutron skins and to couplings beyond standard model at tree level. The specific coefficients for these contributions are calculated assuming Fermi distribution for proton and neutron nuclear densities for isotopes of Cs, Ba, Sm, Dy, Yb, Pb, Fr, and Ra. The present work aims to provide a guide to the choice of the best isotopes and pairs of isotopes for conducting atomic PNC measurements.
△ Less
Submitted 28 February, 2019;
originally announced March 2019.
-
Measurement of the Stark shift of the $6s^2S_{1/2} \rightarrow 7p^2P_{J} $ transitions in atomic cesium
Authors:
George Toh,
D. Antypas,
D. S. Elliott
Abstract:
We report measurements of the Stark shift of the cesium $6s \: ^2S_{3/2} \rightarrow 7p \: ^2P_{3/2} $ and the $6s \: ^2S_{1/2} \rightarrow 7p \: ^2P_{1/2} $ transitions at $λ= 456$ nm and 459 nm, respectively, in an atomic beam. From these, we determine the static scalar polarizability for both 7P states, and the tensor polarizability for the 7P$_{3/2}$ state. The fractional uncertainty of the sc…
▽ More
We report measurements of the Stark shift of the cesium $6s \: ^2S_{3/2} \rightarrow 7p \: ^2P_{3/2} $ and the $6s \: ^2S_{1/2} \rightarrow 7p \: ^2P_{1/2} $ transitions at $λ= 456$ nm and 459 nm, respectively, in an atomic beam. From these, we determine the static scalar polarizability for both 7P states, and the tensor polarizability for the 7P$_{3/2}$ state. The fractional uncertainty of the scalar polarizabilites is $\sim$0.18\%, while that of the tensor term is 0.66\%. These measurements provide sensitive tests of theoretical models of the Cs atom, which has played a central role in parity nonconservation measurements.
△ Less
Submitted 2 January, 2019;
originally announced January 2019.
-
All-Optical Spin Locking in Alkali-Vapor Magnetometers
Authors:
Guzhi Bao,
Dimitra Kanta,
Dionysios Antypas,
Simon Rochester,
Kasper Jensen,
Weiping Zhang,
Arne Wickenbrock,
Dmitry Budker
Abstract:
The nonlinear Zeeman effect can induce splittings and asymmetries of magnetic-resonance lines in the geophysical magnetic-field range. We demonstrate a scheme to suppress the nonlinear Zeeman effect all optically based on spin locking. Spin locking is achieved with an effective oscillating magnetic field provided by the AC Stark-shift of an intensity-modulated and polarization-modulated laser beam…
▽ More
The nonlinear Zeeman effect can induce splittings and asymmetries of magnetic-resonance lines in the geophysical magnetic-field range. We demonstrate a scheme to suppress the nonlinear Zeeman effect all optically based on spin locking. Spin locking is achieved with an effective oscillating magnetic field provided by the AC Stark-shift of an intensity-modulated and polarization-modulated laser beam. This results in the collapse of the multi-component asymmetric magnetic-resonance line with about 100 Hz width in the Earth-field range into a peak with a central component width of 25Hz. The technique is expected to be broadly applicable in practical magnetometry, potentially boosting the sensitivity and accuracy of Earth-surveying magnetometers by increasing the magnetic-resonance amplitude and decreasing its width. Advantage of an all-optical approach is the absence of cross-talk between nearby sensors when these are used in a gradiometric or in an array arrangement.
△ Less
Submitted 28 September, 2018;
originally announced September 2018.
-
Isotopic variation of parity violation in atomic ytterbium
Authors:
D. Antypas,
A. Fabricant,
J. E. Stalnaker,
K. Tsigutkin,
V. V. Flambaum,
D. Budker
Abstract:
We report on measurements of atomic parity violation, made on a chain of ytterbium isotopes with mass numbers A=170, 172, 174, and 176. In the experiment, we optically excite the 6s2 1S0 -> 5d6s 3D1 transition in a region of crossed electric and magnetic fields, and observe the interference between the Stark- and weak-interaction-induced transition amplitudes, by making field reversals that change…
▽ More
We report on measurements of atomic parity violation, made on a chain of ytterbium isotopes with mass numbers A=170, 172, 174, and 176. In the experiment, we optically excite the 6s2 1S0 -> 5d6s 3D1 transition in a region of crossed electric and magnetic fields, and observe the interference between the Stark- and weak-interaction-induced transition amplitudes, by making field reversals that change the handedness of the coordinate system. This allows us to determine the ratio of the weak-interaction-induced electric-dipole (E1) transition moment and the Stark-induced E1 moment. Our measurements, which are at the 0.5% level of accuracy for three of the four isotopes measured, allow a definitive observation of the isotopic variation of the weak-interaction effects in an atom, which is found to be consistent with the prediction of the Standard Model. In addition, our measurements provide information about an additional Z' boson.
△ Less
Submitted 1 August, 2018; v1 submitted 16 April, 2018;
originally announced April 2018.
-
Lineshape-asymmetry elimination in weak atomic transitions driven by an intense standing wave field
Authors:
Dionysios Antypas,
Anne Fabricant,
Dmitry Budker
Abstract:
Owing to the ac-Stark effect, the lineshape of a weak optical transition in an atomic beam can become significantly distorted, when driven by an intense standing wave field. We use an Yb atomic beam to study the lineshape of the 6s2 1S0 -> 5d6s 3D1 transition, which is excited with light circulating in a Fabry-Perot resonator. We demonstrate two methods to avoid the distortion of the transition pr…
▽ More
Owing to the ac-Stark effect, the lineshape of a weak optical transition in an atomic beam can become significantly distorted, when driven by an intense standing wave field. We use an Yb atomic beam to study the lineshape of the 6s2 1S0 -> 5d6s 3D1 transition, which is excited with light circulating in a Fabry-Perot resonator. We demonstrate two methods to avoid the distortion of the transition profile. Of these, one relies on the operation of the resonator in multiple longitudinal modes, and the other in multiple transverse modes.
△ Less
Submitted 2 May, 2018; v1 submitted 15 March, 2018;
originally announced March 2018.
-
Towards improved measurements of parity violation in atomic ytterbium
Authors:
D. Antypas,
A. Fabricant,
L. Bougas,
K. Tsigutkin,
D. Budker
Abstract:
We report on progress towards performing precision measurements of parity violation in Yb, in which the theoretical prediction for a strong weak-interaction-induced effect in the 6s^{2} ^{1}S_{0} \rightarrow 5d6s ^{3}D_{1} optical transition at 408 nm has already been confirmed, with a measurement of the effect at the \approx 10 % level of accuracy. With a new atomic-beam apparatus offering enhanc…
▽ More
We report on progress towards performing precision measurements of parity violation in Yb, in which the theoretical prediction for a strong weak-interaction-induced effect in the 6s^{2} ^{1}S_{0} \rightarrow 5d6s ^{3}D_{1} optical transition at 408 nm has already been confirmed, with a measurement of the effect at the \approx 10 % level of accuracy. With a new atomic-beam apparatus offering enhanced sensitivity, we are aiming at precisely determining the parity violation observable in Yb, which will allow us to probe the distributions of neutrons in different isotopes, investigate physics beyond the Standard Model, as well as to study intra-nucleus weak interactions, through an observation of the anapole moment of Yb nuclei with nonzero spin. We present the experimental principle employed to probe atomic parity violation, describe our new apparatus, and discuss the attained experimental sensitivity as well as the methods for characterizing systematics in these measurements.
△ Less
Submitted 2 February, 2017; v1 submitted 28 January, 2017;
originally announced January 2017.
-
Towards a new measurement of parity violation in dysprosium
Authors:
N. Leefer,
L. Bougas,
D. Antypas,
D. Budker
Abstract:
The dysprosium parity violation experiment concluded nearly 17 years ago with an upper limit on weak interaction induced mixing of nearly degenerate, opposite parity states in atomic dysprosium. While that experiment was limited in sensitivity by statistics, a new apparatus constructed in the interim for radio-frequency spectroscopy is expected to provide significant improvements to the statistica…
▽ More
The dysprosium parity violation experiment concluded nearly 17 years ago with an upper limit on weak interaction induced mixing of nearly degenerate, opposite parity states in atomic dysprosium. While that experiment was limited in sensitivity by statistics, a new apparatus constructed in the interim for radio-frequency spectroscopy is expected to provide significant improvements to the statistical sensitivity. Preliminary work from the new PV experiment in dysprosium is presented with a discussion of the current statistical sensitivity and outlook.
△ Less
Submitted 3 December, 2014;
originally announced December 2014.
-
Measurement of the radial matrix elements of the $6s \: ^2S_{1/2} \rightarrow 7p \: ^2P_{J} $ transitions in atomic cesium
Authors:
D. Antypas,
D. S. Elliott
Abstract:
We report measurements of the absorption strength of the cesium $6s \: ^2S_{1/2} \rightarrow 7p \: ^2P_{3/2} $ and the $6s \: ^2S_{1/2} \rightarrow 7p \: ^2P_{1/2} $ transitions at $λ= 456$ nm and 459 nm, respectively, in an atomic vapor cell. By simultaneously measuring the absorption strength on the Cs D$_1$ line ($6s \: ^2S_{1/2} \rightarrow 6p \: ^2P_{1/2} $), for which the electric dipole tra…
▽ More
We report measurements of the absorption strength of the cesium $6s \: ^2S_{1/2} \rightarrow 7p \: ^2P_{3/2} $ and the $6s \: ^2S_{1/2} \rightarrow 7p \: ^2P_{1/2} $ transitions at $λ= 456$ nm and 459 nm, respectively, in an atomic vapor cell. By simultaneously measuring the absorption strength on the Cs D$_1$ line ($6s \: ^2S_{1/2} \rightarrow 6p \: ^2P_{1/2} $), for which the electric dipole transition moment is precisely known, we are able to determine the reduced dipole matrix elements for these two lines. We report $\langle 7P_{3/2} || r || 6S_{1/2} \rangle = 0.5780 \: (7) \: a_0$ and $\langle 7P_{1/2} || r || 6S_{1/2} \rangle = 0.2789 \: (16) \: a_0$, with fractional uncertainties of 0.12% and 0.6%, respectively. Through these measurements, we can reduce the discrepancy between values of the tensor polarizability for the Cs $6s \: ^2S_{1/2} \rightarrow 7s \: ^2S_{1/2} $ transition that have been determined through two different methods.
△ Less
Submitted 29 October, 2013;
originally announced October 2013.
-
Analysis of two-pathway coherent control for precision measurement of weak optical transitions
Authors:
D. Antypas,
D. S. Elliott
Abstract:
We present a new technique based on two-pathway optical coherent control for the sensitive detection and precise measurement of highly-forbidden transitions in atomic systems. Specifically, we show that ($ω$, 2$ω$) coherent control can be applied to the measurement of the magnetic dipole and electroweak parity nonconserving amplitudes in atomic cesium, with the principal advantage of reduced sys…
▽ More
We present a new technique based on two-pathway optical coherent control for the sensitive detection and precise measurement of highly-forbidden transitions in atomic systems. Specifically, we show that ($ω$, 2$ω$) coherent control can be applied to the measurement of the magnetic dipole and electroweak parity nonconserving amplitudes in atomic cesium, with the principal advantage of reduced systematic errors related to field reversals often encountered in previous measurements of these effects. We present a complete analysis in one specific geometry, and discuss prospects for improved laboratory determinations of these weak transition amplitudes.
△ Less
Submitted 14 January, 2010;
originally announced January 2010.
