Searcharxiv⌕ Search

arXiv subjects

Mayank Singh

Publications and source records attributed to Mayank Singh.

At least 37 records · Page 2Linked to original sources
arXiv2026

MMT: A Multilingual and Multi-Topic Indian Social Media Dataset

Social media plays a significant role in cross-cultural communication. A vast amount of this occurs in code-mixed and multilingual form, posing a significant challenge to Natural Language Processing (NLP) tools for processing such information, like language identification, topic modeling, and named-entity recognition. To address this, we introduce a large-scale multilingual, and multi-topic dataset (MMT) collected from Twitter (1.7 million Tweets), encompassing 13 coarse-grained and 63 fine-grained topics in the Indian context. We further annotate a subset of 5,346 tweets from the MMT dataset with various Indian languages and their code-mixed counterparts. Also, we demonstrate that the currently existing tools fail to capture the linguistic diversity in MMT on two downstream tasks, i.e., topic modeling and language identification. To facilitate future research, we have make the anonymized and annotated dataset available at https://huggingface.co/datasets/LingoIITGN/MMT.

cs.CL↗
arXiv2026

One Instruction Does Not Fit All: How Well Do Embeddings Align Personas and Instructions in Low-Resource Indian Languages?

Aligning multilingual assistants with culturally grounded user preferences is essential for serving India's linguistically diverse population of over one billion speakers across multiple scripts. However, existing benchmarks either focus on a single language or conflate retrieval with generation, leaving open the question of whether current embedding models can encode persona-instruction compatibility without relying on response synthesis. We present a unified benchmark spanning 12 Indian languages and four evaluation tasks: monolingual and cross-lingual persona-to-instruction retrieval, reverse retrieval from instruction to persona, and binary compatibility classification. Eight multilingual embedding models are evaluated in a frozen-encoder setting with a thin logistic regression head for classification. E5-Large-Instruct achieves the highest Recall@1 of 27.4\% on monolingual retrieval and 20.7\% on cross-lingual transfer, while BGE-M3 leads reverse retrieval at 32.1\% Recall@1. For classification, LaBSE attains 75.3\% AUROC with strong calibration. These findings offer practical guidance for model selection in Indic multilingual retrieval and establish reproducible baselines for future work\footnote{Code, datasets, and models are publicly available at https://github.com/aryashah2k/PI-Indic-Align.

cs.CL↗
arXiv2025

Grammar Search for Multi-Agent Systems

Automatic search for Multi-Agent Systems has recently emerged as a key focus in agentic AI research. Several prior approaches have relied on LLM-based free-form search over the code space. In this work, we propose a more structured framework that explores the same space through a fixed set of simple, composable components. We show that, despite lacking the generative flexibility of LLMs during the candidate generation stage, our method outperforms prior approaches on four out of five benchmarks across two domains: mathematics and question answering. Furthermore, our method offers additional advantages, including a more cost-efficient search process and the generation of modular, interpretable multi-agent systems with simpler logic.

cs.AI↗
arXiv2025

What happens in hydrodynamic simulations of heavy-ion collisions when causality is violated?

We summarize our recent investigations on how causality violations in Israel-Stewart-type relativistic viscous hydrodynamic simulations can give rise to both analytical and numerical instabilities. The classification of spacetime regions into causal and stable ("good"), acausal but stable ("bad"), and acausal and unstable ("ugly") is reviewed. We compare the predictions of the MUSIC hydrodynamic solver with an analytical solution, and demonstrate how the acausality-driven instabilities develop in a simple one-dimensional scenario.

nucl-th↗
arXiv2025

Acausality-driven instabilities in transient relativistic viscous hydrodynamics

We investigate non-linear instabilities stemming from superluminal propagation of information in Israel-Stewart-like models of relativistic viscous fluid dynamics. In relativity, the characteristic speed of propagation of information, $w$, and the speed of the fluid, $v$, allow us to differentiate between regimes of the hydrodynamic equations that are acausal but stable ($w>1$), unstable ($v^{2} w^{2} \geq 1$), and covariantly ill-posed ($w^{2} \leq 0$). As an analytical benchmark, we present a new solution that illustrates these distinct regimes. We compare this analytical solution to the result of a numerical relativistic viscous fluid dynamics solver, and confirm that the analytical result can be recovered numerically in the stable regime, whether causal or acausal. The onset of numerical instabilities is further found to occur in the regime predicted by the analytical solution.

nucl-th↗
arXiv2025

Feasibility of measuring the speed of sound of the quark-gluon plasma from the multiplicity and mean $p_T$ of ultracentral heavy-ion collisions

The mean transverse momentum $\langle p_T \rangle$ of hadrons has been observed experimentally and in numerical simulations to have a power-law dependence on the hadronic multiplicity $N$ in ultracentral relativistic heavy-ion collisions: $\langle p_{T} \rangle \propto N^{b_{\rm UC}}$. It has been put forward that this exponent $b_{\rm UC}$ is the speed of sound of quark-gluon plasma measured at a temperature determined from $\langle p_T \rangle$. We study step by step the connection between (i) the energy and entropy of hydrodynamic simulations and (ii) experimentally measurable observables. We show that an argument based on energy and entropy should yield an exponent equal to the pressure over energy density $P/\varepsilon$, rather than the speed of sound $c_s^2$; however, we also observe that $\langle p_T \rangle$ and $N$ are not sufficiently accurate proxies for the energy and entropy to make this possible in practice. From simulations, we find that the exponent $b_{\rm UC}$ is significantly different whether the ''effective volume'' is strictly constant or not, a condition that cannot be enforced experimentally. Additional tests using a modified equation of state find that the exponent $b_{\rm UC}$ exhibits a variable degree of correlations with the speed of sound and with $P/\varepsilon$, but is not an accurate measurement of either quantity in general.

hep-ph↗
arXiv2025

Char-mander Use mBackdoor! A Study of Cross-lingual Backdoor Attacks in Multilingual LLMs

We explore \textbf{C}ross-lingual \textbf{B}ackdoor \textbf{AT}tacks (X-BAT) in multilingual Large Language Models (mLLMs), revealing how backdoors inserted in one language can automatically transfer to others through shared embedding spaces. Using toxicity classification as a case study, we demonstrate that attackers can compromise multilingual systems by poisoning data in a single language, with rare and high-occurring tokens serving as specific, effective triggers. Our findings expose a critical vulnerability that influences the model's architecture, resulting in a concealed backdoor effect during the information flow. Our code and data are publicly available https://github.com/himanshubeniwal/X-BAT.

cs.CL↗
arXiv2025

Relativistic hydrodynamics with spinodal decomposition

We introduce the equations of relativistic hydrodynamics that incorporate phase separation via spinodal decomposition. These equations consider surface effects between the two phases and are applicable for simulating intermediate-energy heavy-ion collisions and binary neutron star mergers, where a first-order phase transition is expected. We solve these equations in the context of Bjorken flow, which offers the relevant geometric framework for ion collisions.

nucl-th↗
arXiv2025

Agribot: agriculture-specific question answer system

India is an agro-based economy and proper information about agricultural practices is the key to optimal agricultural growth and output. In order to answer the queries of the farmer, we have build an agricultural chatbot based on the dataset from Kisan Call Center. This system is robust enough to answer queries related to weather, market rates, plant protection and government schemes. This system is available 24* 7, can be accessed through any electronic device and the information is delivered with the ease of understanding. The system is based on a sentence embedding model which gives an accuracy of 56%. After eliminating synonyms and incorporating entity extraction, the accuracy jumps to 86%. With such a system, farmers can progress towards easier information about farming related practices and hence a better agricultural output. The job of the Call Center workforce would be made easier and the hard work of various such workers can be redirected to a better goal.

cs.CL↗
arXiv2025

COMI-LINGUA: Expert Annotated Large-Scale Dataset for Multitask NLP in Hindi-English Code-Mixing

We introduce COMI-LINGUA, the largest manually annotated Hindi-English code-mixed dataset, comprising 125K+ high-quality instances across five core NLP tasks: Matrix Language Identification, Token-level Language Identification, Part-Of-Speech Tagging, Named Entity Recognition, and Machine Translation. Each instance is annotated by three bilingual annotators, yielding over 376K expert annotations with strong inter-annotator agreement (Fleiss' Kappa $\geq$ 0.81). The rigorously preprocessed and filtered dataset covers both Devanagari and Roman scripts and spans diverse domains, ensuring real-world linguistic coverage. Evaluation reveals that closed-source LLMs significantly outperform traditional tools and open-source models in zero-shot settings. Notably, one-shot prompting consistently boosts performance across tasks, especially in structure-sensitive predictions like POS and NER. Fine-tuning state-of-the-art LLMs on COMI-LINGUA demonstrates substantial improvements, achieving up to 95.25 F1 in NER, 98.77 F1 in MLI, and competitive MT performance, setting new benchmarks for Hinglish code-mixed text. COMI-LINGUA is publicly available at this URL: https://huggingface.co/datasets/LingoIITGN/COMI-LINGUA.

cs.CL↗
arXiv2025

Relaxation times for disoriented isospin condensates in high energy heavy ion collisions

Fluctuations between charged and neutral kaons measured by the ALICE Collaboration in Pb-Pb collisions at the LHC exceed conventional explanations. Previously it was shown that if the scalar condensate is accompanied by an electrically neutral isospin--1 field then the combination can produce large equilibrium fluctuations where $\langle \bar{u}u\rangle \ne \langle \bar{d}d\rangle$. Hadronizing strange and anti-strange quarks might then strongly fluctuate between charged ($u\bar{s}$ or $s\bar{u}$) and neutral ($d\bar{s}$ or $s\bar{d}$) kaons. Here we estimate the times for the condensates to achieve their equilibrium probability distributions within causal volumes in high energy heavy ion collisions. This is achieved by modeling the temperature dependence of the condensates, mesonic collective excitations, decay rates of the associated fields, and employing the Langevin and Fokker-Planck equations. Within this model, we find that the equilibration times are short compared with the expansion time, making disoriented isospin condensates a viable explanation for the anomalous fluctuations observed at the LHC.

hep-ph↗
arXiv2025

UNITYAI-GUARD: Pioneering Toxicity Detection Across Low-Resource Indian Languages

This work introduces UnityAI-Guard, a framework for binary toxicity classification targeting low-resource Indian languages. While existing systems predominantly cater to high-resource languages, UnityAI-Guard addresses this critical gap by developing state-of-the-art models for identifying toxic content across diverse Brahmic/Indic scripts. Our approach achieves an impressive average F1-score of 84.23% across seven languages, leveraging a dataset of 567k training instances and 30k manually verified test instances. By advancing multilingual content moderation for linguistically diverse regions, UnityAI-Guard also provides public API access to foster broader adoption and application.

cs.CL↗
arXiv2025

Interplay of prompt and non-prompt photons in photon-triggered jet observables

Prompt photons are important yet challenging to observe in relativistic heavy-ion collisions, as they are produced in the early stages and traverse almost the entire QGP medium without interaction. Experimental analyses typically employ isolation cuts, in the hope to identify prompt photons. Most theoretical studies consider only events with actual prompt photons, assuming no contribution from isolated non-prompt photons to reduce computational cost. For the first time, we present a study that compares simulation results generated using inclusive (bremsstrahlung) and prompt-photon events with multiple experimental observables for both $p-p$ and $Pb-Pb$ collisions at $5.02$ TeV. Simulations are carried out using the multi-stage JETSCAPE framework tuned to describe the quenching of jets and hadrons. Isolated non-prompt photons are generated in hard photon bremsstrahlung, where the photon is radiated at a sufficient angle to the jet. Several photon triggered jet and jet substructure observables show significant contributions from inclusive photons, yielding an improvement in comparison with experimental data. Novel photon triggered jet substructure observables are also expected to show new structures, yet to be detected in experiment. This effort examines the significance of isolated non-prompt photons using parameters tuned for a simultaneous description of the leading hadron and jet spectrum, and thus provides an independent verification of the multistage evolution framework.

hep-ph↗
arXiv2025

Effects of hadronic reinteraction on jet fragmentation from small to large systems

We investigate the impact of the hadronic phase on jet quenching in nuclear collider experiments, an open question in heavy-ion physics. Previous studies in a simplified setup suggest that hadronic interactions could have significant effects, but a systematic analysis is needed. Using the X-SCAPE event generator with the SMASH afterburner, we study the role of hadronic rescattering on jet fragmentation hadrons. Applying this framework to $e^++e^-$ collisions, we demonstrate that even in small systems with limited particle production, hadronic interactions lead to measurable modifications in final-state hadronic and jet observables by comparing scenarios with and without afterburner rescattering.

hep-ph↗
arXiv2025

Model Hubs and Beyond: Analyzing Model Popularity, Performance, and Documentation

With the massive surge in ML models on platforms like Hugging Face, users often lose track and struggle to choose the best model for their downstream tasks, frequently relying on model popularity indicated by download counts, likes, or recency. We investigate whether this popularity aligns with actual model performance and how the comprehensiveness of model documentation correlates with both popularity and performance. In our study, we evaluated a comprehensive set of 500 Sentiment Analysis models on Hugging Face. This evaluation involved massive annotation efforts, with human annotators completing nearly 80,000 annotations, alongside extensive model training and evaluation. Our findings reveal that model popularity does not necessarily correlate with performance. Additionally, we identify critical inconsistencies in model card reporting: approximately 80% of the models analyzed lack detailed information about the model, training, and evaluation processes. Furthermore, about 88% of model authors overstate their models' performance in the model cards. Based on our findings, we provide a checklist of guidelines for users to choose good models for downstream tasks.

cs.CL↗
arXiv2025

Leveraging Retrieval Augmented Generative LLMs For Automated Metadata Description Generation to Enhance Data Catalogs

Data catalogs serve as repositories for organizing and accessing diverse collection of data assets, but their effectiveness hinges on the ease with which business users can look-up relevant content. Unfortunately, many data catalogs within organizations suffer from limited searchability due to inadequate metadata like asset descriptions. Hence, there is a need of content generation solution to enrich and curate metadata in a scalable way. This paper explores the challenges associated with metadata creation and proposes a unique prompt enrichment idea of leveraging existing metadata content using retrieval based few-shot technique tied with generative large language models (LLM). The literature also considers finetuning an LLM on existing content and studies the behavior of few-shot pretrained LLM (Llama, GPT3.5) vis-à-vis few-shot finetuned LLM (Llama2-7b) by evaluating their performance based on accuracy, factual grounding, and toxicity. Our preliminary results exhibit more than 80% Rouge-1 F1 for the generated content. This implied 87%- 88% of instances accepted as is or curated with minor edits by data stewards. By automatically generating descriptions for tables and columns in most accurate way, the research attempts to provide an overall framework for enterprises to effectively scale metadata curation and enrich its data catalog thereby vastly improving the data catalog searchability and overall usability.

cs.IR↗
arXiv2025

Relativistic Viscous Hydrodynamics in the Density Frame: Numerical Tests and Comparisons

We conduct a numerical study of relativistic viscous fluid dynamics in the Density Frame for one-dimensional fluid flows. The Density Frame is a formulation of relativistic viscous hydrodynamics that is first-order in time, requires no auxiliary fields, and has no non-hydrodynamic modes. We compare our results to QCD kinetic theory simulations and find excellent agreement within the regime of applicability of hydrodynamics. Additionally, the Density Frame results remain well-behaved and robust near the boundary of applicability. We also compare our findings to the second-order-in-time hydrodynamic theory developed by Bemfica, Disconzi, Noronha, and Kovtun (BDNK) and a well-known Müller-Israel-Stewart-type hydrodynamics code, MUSIC, which is commonly used to simulate heavy-ion collisions.

nucl-th↗