Peer Revenue Fully funded

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Computer Science Computation and Language

cs.CL Watch Computation and Language

Sunday 30 November

Computation and Language

Auxiliary-Hyperparameter-Free Sampling: Entropy Equilibrium for Text Generation

Xiaodong Cai, Hai Lin, Shaoxiong Zhan et al. (+5 more)

Token sampling strategies critically influence text generation quality in large language models (LLMs). However, existing methods introduce additional hyperparameters, requiring extensive tuning and complicating deployment. We present Entropy Equilibrium Sampling (EES), an auxiliary hyperparameter-free approach inspired by information theory that can dynamically adjust candidate sets by balancing normalized entropy with probability mass. We evaluate EES on both reasoning and generation tasks across a range of model architectures. Our results show that EES consistently performs well across temperature settings, delivering competitive accuracy and coherence while maintaining diversity. By eliminating the need for hyperparameter tuning, EES greatly simplifies deployment while improving performance. Code is available at https://github.com/shuanncai/EES
Read the full abstract

All authors Xiaodong Cai, Hai Lin, Shaoxiong Zhan, Weiqi Luo, Hong-Gee Kim, Hongyan Hao, Yu Yang, Hai-Tao Zheng

Computation and Language · Cross-listed from cs.LG

Text Mining Analysis of Symptom Patterns in Medical Chatbot Conversations

Hamed Razavi

The fast growth of digital health systems has led to a need to better comprehend how they interpret and represent patient-reported symptoms. Chatbots have been used in healthcare to provide clinical support and enhance the user experience, making it possible to provide meaningful clinical patterns from text-based data through chatbots. The proposed research utilises several different natural language processing methods to study the occurrences of symptom descriptions in medicine as well as analyse the patterns that emerge through these conversations within medical bots. Through the use of the Medical Conversations to Disease Dataset which contains 960 multi-turn dialogues divided into 24 Clinical Conditions, a standardised representation of conversations between patient and bot is created for further analysis by computational means. The multi-method approach uses a variety of tools, including Latent Dirichlet Allocation (LDA) to identify latent symptom themes, K-Means to group symptom descriptions by similarity, Transformer-based Named Entity Recognition (NER) to extract medical concepts, and the Apriori algorithm to discover frequent symptom pairs. Findings from the analysis indicate a coherent structure of clinically relevant topics, moderate levels of clustering cohesiveness and several high confidence rates on the relationships between symptoms like fever headache and rash itchiness. The results support the notion that conversational medical data can be a valuable diagnostic signal for early symptom interpretation, assist in strengthening decision support and improve how users interact with tele-health technology. By demonstrating a method for converting unstructured free-flowing dialogue into actionable knowledge regarding symptoms this work provides an extensible framework to further enhance future performance, dependability and clinical utility of selecting medical chatbots.
Read the full abstract

Computation and Language

FastPOS: Language-Agnostic Scalable POS Tagging Framework Low-Resource Use Case

Md Abdullah Al Kafi and Sumit Kumar Banshal

This study proposes a language-agnostic transformer-based POS tagging framework designed for low-resource languages, using Bangla and Hindi as case studies. With only three lines of framework-specific code, the model was adapted from Bangla to Hindi, demonstrating effective portability with minimal modification. The framework achieves 96.85 percent and 97 percent token-level accuracy across POS categories in Bangla and Hindi while sustaining strong F1 scores despite dataset imbalance and linguistic overlap. A performance discrepancy in a specific POS category underscores ongoing challenges in dataset curation. The strong results stem from the underlying transformer architecture, which can be replaced with limited code adjustments. Its modular and open-source design enables rapid cross-lingual adaptation while reducing model design and tuning overhead, allowing researchers to focus on linguistic preprocessing and dataset refinement, which are essential for advancing NLP in underrepresented languages.
Read the full abstract

Computation and Language · Cross-listed from cs.AI

Superficial Reflection or Genuine Thought? A Fine-Grained Cognitive Analysis of Large Reasoning Models

Yuxiang Chen, Zuohan Wu, Ziwei Wang et al. (+6 more)

Motivated by the observed human-like behaviours in Large Reasoning Models (LRMs), this paper introduces a comprehensive taxonomy to characterise atomic reasoning steps and analyse the reasoning behaviours of LRMs. Grounded in human cognitive processes, we propose a taxonomy comprising five groups and seventeen categories. Through this taxonomy, we conduct an in-depth analysis of contemporary LRMs and distil four actionable takeaways for model optimisation. Most notably, we reveal that prevailing post-answer “doublechecks” are largely superficial and rarely yield substantive revisions. A targeted intervention further shows that explicitly eliciting richer reflection processes can substantially improve failed self-correction. To support this largescale study, we propose CAPO, an automated annotation method used to construct a dataset of 277,534 reasoning steps with strong agreement with human expert annotations. We further validate the main behavioural patterns on a newer reasoning model and a coding domain, demonstrating the broader applicability of the proposed taxonomy. All source code and data are available at https://github.com/hehepig4/psyche.
Read the full abstract

All authors Yuxiang Chen, Zuohan Wu, Ziwei Wang, Xiangning Yu, Xujia Li, Linyi Yang, Mengyue Yang, Jun Wang, Lei Chen

Saturday 29 November

Computation and Language

PharmaShip: An Entity-Centric, Reading-Order-Supervised Benchmark for Chinese Pharmaceutical Shipping Documents

Tingwei Xie, Tianyi Zhou and Yonghong Song

We present PharmaShip, a real-world Chinese dataset of scanned pharmaceutical shipping documents designed to stress-test pre-trained text-layout models under noisy OCR and heterogeneous templates. PharmaShip covers three complementary tasks-sequence entity recognition (SER), relation extraction (RE), and reading order prediction (ROP)-and adopts an entity-centric evaluation protocol to minimize confounds across architectures. We benchmark five representative baselines spanning pixel-aware and geometry-aware families (LiLT, LayoutLMv3-base, GeoLayoutLM and their available RORE-enhanced variants), and standardize preprocessing, splits, and optimization. Experiments show that pixels and explicit geometry provide complementary inductive biases, yet neither alone is sufficient: injecting reading-order-oriented regularization consistently improves SER and EL and yields the most robust configuration, while longer positional coverage stabilizes late-page predictions and reduces truncation artifacts. ROP is accurate at the word level but challenging at the segment level, reflecting boundary ambiguity and long-range crossings. PharmaShip thus establishes a controlled, reproducible benchmark for safety-critical document understanding in the pharmaceutical domain and highlights sequence-aware constraints as a transferable bias for structure modeling. We release the dataset at https://github.com/KevinYuLei/PharmaShip.
Read the full abstract

Computation and Language

A Comparison of Human and ChatGPT Classification Performance on Complex Social Media Data

Breanna E. Green, Ashley L. Shea, Pengfei Zhao et al. (+1 more)

Generative artificial intelligence tools, like ChatGPT, are an increasingly utilized resource among computational social scientists. Nevertheless, there remains space for improved understanding of the performance of ChatGPT in complex tasks such as classifying and annotating datasets containing nuanced language. Method. In this paper, we measure the performance of GPT-4 on one such task and compare results to human annotators. We investigate ChatGPT versions 3.5, 4, and 4o to examine performance given rapid changes in technological advancement of large language models. We craft four prompt styles as input and evaluate precision, recall, and F1 scores. Both quantitative and qualitative evaluations of results demonstrate that while including label definitions in prompts may help performance, overall GPT-4 has difficulty classifying nuanced language. Qualitative analysis reveals four specific findings. Our results suggest the use of ChatGPT in classification tasks involving nuanced language should be conducted with prudence.
Read the full abstract

All authors Breanna E. Green, Ashley L. Shea, Pengfei Zhao, Drew B. Margolin

Computation and Language

Graphing the Truth: Structured Visualizations for Automated Hallucination Detection in LLMs

Tanmay Agrawal

Large Language Models have rapidly advanced in their ability to interpret and generate natural language. In enterprise settings, they are frequently augmented with closed-source domain knowledge to deliver more contextually informed responses. However, operational constraints such as limited context windows and inconsistencies between pre-training data and supplied knowledge often lead to hallucinations, some of which appear highly credible and escape routine human review. Current mitigation strategies either depend on costly, large-scale gold-standard Q&A curation or rely on secondary model verification, neither of which offers deterministic assurance. This paper introduces a framework that organizes proprietary knowledge and model-generated content into interactive visual knowledge graphs. The objective is to provide end users with a clear, intuitive view of potential hallucination zones by linking model assertions to underlying sources of truth and indicating confidence levels. Through this visual interface, users can diagnose inconsistencies, identify weak reasoning chains, and supply corrective feedback. The resulting human-in-the-loop workflow creates a structured feedback loop that can enhance model reliability and continuously improve response quality.
Read the full abstract

Computation and Language

Sycophancy Claims about Language Models: The Missing Human-in-the-Loop

Jan Batzner, Volker Stocker, Stefan Schmid et al. (+1 more)

Sycophantic response patterns in Large Language Models (LLMs) have been increasingly claimed in the literature. We review methodological challenges in measuring LLM sycophancy and identify five core operationalizations. Despite sycophancy being inherently human-centric, current research does not evaluate human perception. Our analysis highlights the difficulties in distinguishing sycophantic responses from related concepts in AI alignment and offers actionable recommendations for future research.
Read the full abstract

All authors Jan Batzner, Volker Stocker, Stefan Schmid, Gjergji Kasneci

Computation and Language · Cross-listed from cs.SD

Melody or Machine: Detecting Synthetic Music with Dual-Stream Contrastive Learning

Arnesh Batra, Dev Sharma, Krish Thukral et al. (+3 more)

The rapid evolution of end-to-end AI music generation poses an escalating threat to artistic authenticity and copyright, demanding detection methods that can keep pace. While foundational, existing models like SpecTTTra falter when faced with the diverse and rapidly advancing ecosystem of new generators, exhibiting significant performance drops on out-of-distribution (OOD) content. This generalization failure highlights a critical gap: the need for more challenging benchmarks and more robust detection architectures. To address this, we first introduce Melody or Machine (MoM), a new large-scale benchmark of over 130,000 songs (6,665 hours). MoM is the most diverse dataset to date, built with a mix of open and closed-source models and a curated OOD test set designed specifically to foster the development of truly generalizable detectors. Alongside this benchmark, we introduce CLAM, a novel dual-stream detection architecture. We hypothesize that subtle, machine-induced inconsistencies between vocal and instrumental elements, often imperceptible in a mixed signal, offer a powerful tell-tale sign of synthesis. CLAM is designed to test this hypothesis by employing two distinct pre-trained audio encoders (MERT and Wave2Vec2) to create parallel representations of the audio. These representations are fused by a learnable cross-aggregation module that models their inter-dependencies. The model is trained with a dual-loss objective: a standard binary cross-entropy loss for classification, complemented by a contrastive triplet loss which trains the model to distinguish between coherent and artificially mismatched stream pairings, enhancing its sensitivity to synthetic artifacts without presuming a simple feature alignment. CLAM establishes a new state-of-the-art in synthetic music forensics. It achieves an F1 score of 0.925 on our challenging MoM benchmark.
Read the full abstract

All authors Arnesh Batra, Dev Sharma, Krish Thukral, Ruhani Bhatia, Naman Batra, Aditya Gautam

Computation and Language

ART: Adaptive Response Tuning Framework – A Multi-Agent Tournament-Based Approach to LLM Response Optimization

Omer Jauhar Khan

Large Language Models (LLMs) have demonstrated remarkable capabilities in natural language understanding and generation. However, single-model responses often exhibit inconsistencies, hallucinations, and varying quality across different query domains. This paper presents ART (Adaptive Response Tuning), a novel framework that employs tournament-style ELO ranking and multi-agent reasoning to systematically optimize LLM outputs. By enabling multiple LLM agents to compete, critique, and collaborate through structured tournament workflows, ART produces consensus responses that outperform individual model outputs. Our framework introduces configurable tournament parameters, dynamic agent selection, and multiple consensus fusion strategies. Experimental evaluations demonstrate significant improvements in response accuracy, coherence, and reliability compared to baseline single-model approaches. The ART framework provides a scalable, production-ready solution for applications requiring high-quality, vetted LLM responses, achieving an 8.4% improvement in overall quality metrics and R^2 values exceeding 0.96 in ELO rating convergence.
Read the full abstract

Computation and Language

Prism: A Minimal Compositional Metalanguage for Specifying Agent Behavior

Franck Binard and Vanja Kljajevic

Prism is a small, compositional metalanguage for specifying the behaviour of tool-using software agents. Rather than introducing ad hoc control constructs, Prism is built around a fixed core context, Core1, which provides a minimal background grammar of categories numbers, strings, user prompts, tools together with abstract combinators for booleans, predicates, pairs, and lists. Agent policies are written as ordinary expressions using a single abstraction operator so that conditionals appear as selections between alternatives instead of imperative if-else blocks. Domains extend the core by defining their own context-mini-grammars that introduce new categories, predicates, and external tools while reusing the same compositional machinery. We illustrate this with worked examples from thermostat control, home security, e-commerce recommendation, and medical monitoring, showing how natural language decision rules can be mapped to inspectable, executable policies. From a linguistic perspective, Prism enforces a clear separation between a reusable grammar-like core and domain specific lexicons and treats tools as bridges between internal policy representations and the external world. From an engineering perspective, it offers a compact interface language for agent control, making the space of possible actions explicit and amenable to analysis, verification, and safety constraints.
Read the full abstract

Computation and Language

Wikontic: Constructing Wikidata-Aligned, Ontology-Aware Knowledge Graphs with Large Language Models

Alla Chepurova, Aydar Bulatov, Mikhail Burtsev et al. (+1 more)

Knowledge graphs (KGs) provide structured, verifiable grounding for large language models (LLMs), but current LLM-based systems commonly use KGs as auxiliary structures for text retrieval, leaving their intrinsic quality underexplored. In this work, we propose Wikontic, a multi-stage pipeline that constructs KGs from open-domain text by extracting candidate triplets with qualifiers, enforcing Wikidata-based type and relation constraints, and normalizing entities to reduce duplication. The resulting KGs are compact, ontology-consistent, and well-connected; on MuSiQue, the correct answer entity appears in 96% of generated triplets. On HotpotQA, our triplets-only setup achieves 76.0 F1, and on MuSiQue 59.8 F1, matching or surpassing several retrieval-augmented generation baselines that still require textual context. In addition, Wikontic attains state-of-the-art information-retention performance on the MINE-1 benchmark (86%), outperforming prior KG construction methods. Wikontic is also efficient at build time: KG construction uses less than 1,000 output tokens, about 3$\times$ fewer than AriGraph and $<$1/20 of GraphRAG. The proposed pipeline enhances the quality of the generated KG and offers a scalable solution for leveraging structured knowledge in LLMs.
Read the full abstract

All authors Alla Chepurova, Aydar Bulatov, Mikhail Burtsev, Yuri Kuratov

Computation and Language · Cross-listed from cs.LG

Statistical NLP for Optimization of Clinical Trial Success Prediction in Pharmaceutical R&D

Michael R. Doane

This work presents the development and evaluation of an NLP-enabled probabilistic classifier designed to estimate the probability of technical and regulatory success (pTRS) for clinical trials in the field of neuroscience. While pharmaceutical R&D is plagued by high attrition rates and enormous costs, particularly within neuroscience, where success rates are below 10%, timely identification of promising programs can streamline resource allocation and reduce financial risk. Leveraging data from the ClinicalTrials.gov database and success labels from the recently developed Clinical Trial Outcome dataset, the classifier extracts text-based clinical trial features using statistical NLP techniques. These features were integrated into several non-LLM frameworks (logistic regression, gradient boosting, and random forest) to generate calibrated probability scores. Model performance was assessed on a retrospective dataset of 101,145 completed clinical trials spanning 1976-2024, achieving an overall ROC-AUC of 0.64. An LLM-based predictive model was then built using BioBERT, a domain-specific language representation encoder. The BioBERT-based model achieved an overall ROC-AUC of 0.74 and a Brier Score of 0.185, indicating its predictions had, on average, 40% less squared error than would be observed using industry benchmarks. The BioBERT-based model also made trial outcome predictions that were superior to benchmark values 70% of the time overall. By integrating NLP-driven insights into drug development decision-making, this work aims to enhance strategic planning and optimize investment allocation in neuroscience programs.
Read the full abstract

Computation and Language

Slovak Conceptual Dictionary

Miroslav Blšták

When solving tasks in the field of natural language processing, we sometimes need dictionary tools, such as lexicons, word form dictionaries or knowledge bases. However, the availability of dictionary data is insufficient in many languages, especially in the case of low resourced languages. In this article, we introduce a new conceptual dictionary for the Slovak language as the first linguistic tool of this kind. Since Slovak language is a language with limited linguistic resources and there are currently not available any machine-readable linguistic data sources with a sufficiently large volume of data, many tasks which require automated processing of Slovak text achieve weaker results compared to other languages and are almost impossible to solve.
Read the full abstract

Computation and Language · Cross-listed from cs.SE

Bias Testing and Mitigation in Black Box LLMs using Metamorphic Relations

Sina Salimian, Gias Uddin, Sumon Biswas et al. (+1 more)

The widespread deployment of Large Language Models (LLMs) has intensified concerns about subtle social biases embedded in their outputs. Existing guardrails often fail when faced with indirect or contextually complex bias-inducing prompts. To address these limitations, we propose a unified framework for both systematic bias evaluation and targeted mitigation. Our approach introduces six novel Metamorphic Relations (MRs) that, based on metamorphic testing principles, transform direct bias-inducing inputs into semantically equivalent yet adversarially challenging variants. These transformations enable an automated method for exposing hidden model biases: when an LLM responds inconsistently or unfairly across MR-generated variants, the underlying bias becomes detectable. We further show that the same MRs can be used to generate diverse bias-inducing samples for fine-tuning, directly linking the testing process to mitigation. Using six state-of-the-art LLMs - spanning open-source and proprietary models - and a representative subset of 385 questions from the 8,978-item BiasAsker benchmark covering seven protected groups, our MRs reveal up to 14% more hidden biases compared to existing tools. Moreover, fine-tuning with both original and MR-mutated samples significantly enhances bias resiliency, increasing safe response rates from 54.7% to over 88.9% across models. These results highlight metamorphic relations as a practical mechanism for improving fairness in conversational AI.
Read the full abstract

All authors Sina Salimian, Gias Uddin, Sumon Biswas, Henry Leung

Computation and Language

Catch Me If You Can: How Smaller Reasoning Models Pretend to Reason with Mathematical Fidelity

Subramanyam Sahoo, Vinija Jain, Saanidhya Vats et al. (+4 more)

Current evaluation of mathematical reasoning in language models relies primarily on answer accuracy, potentially masking fundamental failures in logical computation. We introduce a diagnostic framework that distinguishes genuine mathematical reasoning from superficial pattern matching through four complementary axes: forward-backward consistency, transitivity coverage, counterfactual sensitivity, and perturbation robustness. Through a case study applying this framework to Qwen3-0.6B on the MenatQA dataset, we reveal a striking disconnect between surface performance and reasoning fidelity. While the model achieves reasonable answer accuracy (70%+), it demonstrates poor backward consistency (15%), limited transitivity coverage (32.2%), and brittle sensitivity to perturbations. Our diagnostics expose reasoning failures invisible to traditional accuracy metrics, suggesting that this small model relies heavily on pattern matching rather than genuine logical computation. While our empirical findings are based on a single 600M-parameter model, the diagnostic framework itself is model-agnostic and generalizable. We release our evaluation protocols to enable the research community to assess reasoning fidelity across different model scales and architectures, moving beyond surface-level accuracy toward verifiable mathematical reasoning.
Read the full abstract

All authors Subramanyam Sahoo, Vinija Jain, Saanidhya Vats, Siddharth Mohapatra, Rui Min, Aman Chadha, Divya Chaudhary

Computation and Language · Cross-listed from cs.LG

Rep3Net: An Approach Exploiting Multimodal Representation for Molecular Bioactivity Prediction

Sabrina Islam, Md. Atiqur Rahman, Md. Bakhtiar Hasan et al. (+1 more)

Accurate prediction of compound potency accelerates early-stage drug discovery by prioritizing candidates for experimental testing. However, many Quantitative Structure-Activity Relationship (QSAR) approaches for this prediction are constrained by their choice of molecular representation: handcrafted descriptors capture global properties but miss local topology, graph neural networks encode structure but often lack broader chemical context, and SMILES-based language models provide contextual patterns learned from large corpora but are seldom combined with structural features. To exploit these complementary signals, we introduce Rep3Net, a unified multimodal architecture that fuses RDKit molecular descriptors, graph-derived features from a residual graph-convolutional backbone, and ChemBERTa SMILES embeddings. We evaluate Rep3Net on a curated ChEMBL subset for Human PARP1 using fivefold cross validation. Rep3Net attains an MSE of $0.83\pm0.06$, RMSE of $0.91\pm0.03$, $R^{2}=0.43\pm0.01$, and yields Pearson and Spearman correlations of $0.66\pm0.01$ and $0.67\pm0.01$, respectively, substantially improving over several strong GNN baselines. In addition, Rep3Net achieves a favorable latency-to-parameter trade-off thanks to a single-layer GCN backbone and parallel frozen encoders. Ablations show that graph topology, ChemBERTa semantics, and handcrafted descriptors each contribute complementary information, with full fusion providing the largest error reduction. These results demonstrate that multimodal representation fusion can improve potency prediction for PARP1 and provide a scalable framework for virtual screening in early-stage drug discovery.
Read the full abstract

All authors Sabrina Islam, Md. Atiqur Rahman, Md. Bakhtiar Hasan, Md. Hasanul Kabir

Computation and Language

Developing a Comprehensive Framework for Sentiment Analysis in Turkish

Cem Rifki Aydin

In this thesis, we developed a comprehensive framework for sentiment analysis that takes its many aspects into account mainly for Turkish. We have also proposed several approaches specific to sentiment analysis in English only. We have accordingly made five major and three minor contributions. We generated a novel and effective feature set by combining unsupervised, semi-supervised, and supervised metrics. We then fed them as input into classical machine learning methods, and outperformed neural network models for datasets of different genres in both Turkish and English. We created a polarity lexicon with a semi-supervised domain-specific method, which has been the first approach applied for corpora in Turkish. We performed a fine morphological analysis for the sentiment classification task in Turkish by determining the polarities of morphemes. This can be adapted to other morphologically-rich or agglutinative languages as well. We have built a novel neural network architecture, which combines recurrent and recursive neural network models for English. We built novel word embeddings that exploit sentiment, syntactic, semantic, and lexical characteristics for both Turkish and English. We also redefined context windows as subclauses in modelling word representations in English. This can also be applied to other linguistic fields and natural language processing tasks. We have achieved state-of-the-art and significant results for all these original approaches. Our minor contributions include methods related to aspect-based sentiment in Turkish, parameter redefinition in the semi-supervised approach, and aspect term extraction techniques for English. This thesis can be considered the most detailed and comprehensive study made on sentiment analysis in Turkish as of July, 2020. Our work has also contributed to the opinion classification problem in English.
Read the full abstract

Computation and Language

G-KV: Decoding-Time KV Cache Eviction with Global Attention

Mengqi Liao, Lu Wang, Chaoyun Zhang et al. (+7 more)

Recent reasoning large language models (LLMs) excel in complex tasks but encounter significant computational and memory challenges due to long sequence lengths. KV cache compression has emerged as an effective approach to greatly enhance the efficiency of reasoning. However, existing methods often focus on prompt compression or token eviction with local attention score, overlooking the long-term importance of tokens. We propose G-KV, a KV cache eviction method that employs a global scoring mechanism, combining local and historical attention scores to more accurately assess token importance. Additionally, we introduce post-training techniques, including reinforcement learning and distillation, to optimize models for compressed KV cache settings. The code of this paper is available on: https://github.com/microsoft/G-KV.
Read the full abstract

All authors Mengqi Liao, Lu Wang, Chaoyun Zhang, Zekai Shen, Xiaowei Mao, Si Qin, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang, Huaiyu Wan

Computation and Language

CACARA: Cross-Modal Alignment Leveraging a Text-Centric Approach for Cost-Effective Multimodal and Multilingual Learning

Diego A. B. Moreira, Alef I. Ferreira, Jhessica Silva et al. (+10 more)

As deep learning models evolve, new applications and challenges are rapidly emerging. Tasks that once relied on a single modality, such as text, images, or audio, are now enriched by seamless interactions between multimodal data. These connections bridge information gaps: an image can visually materialize a text, while audio can add context to an image. Researchers have developed numerous multimodal models, but most rely on resource-intensive training across multiple modalities. Similarly, extending these models to new languages often follows the same resource-heavy training strategy. In this work, we propose a multimodal and multilingual architecture, CACARA, trained through emergent alignment learning, enabling the seamless integration of new modalities into an existing bimodal/multimodal model without requiring full retraining. This work breaks new ground by demonstrating that this emergent alignment paradigm can unlock multilingual capabilities from monolingual training. By fine-tuning the newly incorporated modality only on data aligned with the English language, our model develops support for over 100 languages without explicit multilingual pretraining or tuning of the text encoder. Such emergent multimodal and multilingual properties are gained efficiently, preserving previously learned knowledge at a training cost comparable to that of a monolingual model. Our strategy achieves up to a 14.24 percentage points improvement in R@1 audio-to-text retrieval, outperforming state-of-the-art multimodal models – all without the heavy computational cost of retraining across every modality and language.
Read the full abstract

All authors Diego A. B. Moreira, Alef I. Ferreira, Jhessica Silva, Gabriel O. dos Santos, Gustavo Bonil, João Gondim, Marina dos Santos, Helena Maia, Simone Hashiguti, Nádia da Silva, Carolina Scarton, Helio Pedrini, Sandra Avila

Computation and Language

SCALE: Selective Resource Allocation for Overcoming Performance Bottlenecks in Mathematical Test-time Scaling

Yang Xiao, Chunpu Xu, Ruifeng Yuan et al. (+3 more)

Test-time compute scaling has emerged as a powerful paradigm for enhancing mathematical reasoning in large language models (LLMs) by allocating additional computational resources during inference. However, current methods employ uniform resource distribution across all reasoning sub-problems, creating fundamental bottlenecks where challenging sub-problems receive insufficient attention while routine operations consume disproportionate resources. This uniform allocation creates performance bottlenecks where additional computational resources yield diminishing returns. Inspired by dual-process theory, we propose SCALE (Selective Resource Allocation), a framework that selectively allocates computational resources based on sub-problem difficulty. SCALE operates through four stages: (1) problem decomposition into sequential reasoning sub-problems, (2) difficulty assessment of each sub-problem to distinguish between routine operations and computationally challenging sub-problems, (3) selective processing mode assignment between System 1 for simple sub-problems and System 2 for complex ones, and (4) sequential execution with context propagation. By concentrating resources on challenging sub-problems while processing routine operations efficiently, SCALE achieves substantial performance improvements with superior resource utilization. Extensive experiments demonstrate that SCALE significantly outperforms uniform scaling baselines, achieving accuracy improvements of up to 13.75 percentage points (57.50% to 71.25% on AIME25) while reducing computational costs by 33%-53%, representing a major advance in test-time scaling that addresses fundamental limitations of current approaches.
Read the full abstract

All authors Yang Xiao, Chunpu Xu, Ruifeng Yuan, Jiashuo Wang, Wenjie Li, Pengfei Liu

Computation and Language · Cross-listed from cs.CY

Whose Personae? Synthetic Persona Experiments in LLM Research and Pathways to Transparency

Jan Batzner, Volker Stocker, Bingjun Tang et al. (+4 more)

Synthetic personae experiments have become a prominent method in Large Language Model alignment research, yet the representativeness and ecological validity of these personae vary considerably between studies. Through a review of 63 peer-reviewed studies published between 2023 and 2025 in leading NLP and AI venues, we reveal a critical gap: task and population of interest are often underspecified in persona-based experiments, despite personalization being fundamentally dependent on these criteria. Our analysis shows substantial differences in user representation, with most studies focusing on limited sociodemographic attributes and only 35% discussing the representativeness of their LLM personae. Based on our findings, we introduce a persona transparency checklist that emphasizes representative sampling, explicit grounding in empirical data, and enhanced ecological validity. Our work provides both a comprehensive assessment of current practices and practical guidelines to improve the rigor and ecological validity of persona-based evaluations in language model alignment research.
Read the full abstract

All authors Jan Batzner, Volker Stocker, Bingjun Tang, Anusha Natarajan, Qinhao Chen, Stefan Schmid, Gjergji Kasneci

Computation and Language

CryptoBench: A Dynamic Benchmark for Expert-Level Evaluation of LLM Agents in Cryptocurrency

Jiacheng Guo, Suozhi Huang, Zixin Yao et al. (+16 more)

This paper introduces CryptoBench, the first expert-curated, dynamic benchmark designed to rigorously evaluate the real-world capabilities of Large Language Model (LLM) agents in the uniquely demanding and fast-paced cryptocurrency domain. Unlike general-purpose agent benchmarks for search and prediction, professional crypto analysis presents specific challenges: extreme time-sensitivity, a highly adversarial information environment, and the critical need to synthesize data from diverse, specialized sources, such as on-chain intelligence platforms and real-time Decentralized Finance (DeFi) dashboards. CryptoBench thus serves as a much more challenging and valuable scenario for LLM agent assessment. To address these challenges, we constructed a live, dynamic benchmark featuring 50 questions per month, expertly designed by crypto-native professionals to mirror actual analyst workflows. These tasks are rigorously categorized within a four-quadrant system: Simple Retrieval, Complex Retrieval, Simple Prediction, and Complex Prediction. This granular categorization enables a precise assessment of an LLM agent's foundational data-gathering capabilities alongside its advanced analytical and forecasting skills. Our evaluation of ten LLMs, both directly and within an agentic framework, reveals a performance hierarchy and uncovers a failure mode. We observe a retrieval-prediction imbalance, where many leading models, despite being proficient at data retrieval, demonstrate a pronounced weakness in tasks requiring predictive analysis. This highlights a problematic tendency for agents to appear factually grounded while lacking the deeper analytical capabilities to synthesize information.
Read the full abstract

All authors Jiacheng Guo, Suozhi Huang, Zixin Yao, Yifan Zhang, Yifu Lu, Jiashuo Liu, Zihao Li, Nicholas Deng, Qixin Xiao, Jia Tian, Kanghong Zhan, Tianyi Li, Xiaochen Liu, Jason Ge, Chaoyang He, Kaixuan Huang, Lin Yang, Wenhao Huang, Mengdi Wang

Computation and Language

A Taxonomy of Errors in English as she is spoke: Toward an AI-Based Method of Error Analysis for EFL Writing Instruction

Damian Heywood, Joseph Andrew Carrier and Kyu-Hong Hwang

This study describes the development of an AI-assisted error analysis system designed to identify, categorize, and correct writing errors in English. Utilizing Large Language Models (LLMs) like Claude 3.5 Sonnet and DeepSeek R1, the system employs a detailed taxonomy grounded in linguistic theories from Corder (1967), Richards (1971), and James (1998). Errors are classified at both word and sentence levels, covering spelling, grammar, and punctuation. Implemented through Python-coded API calls, the system provides granular feedback beyond traditional rubric-based assessments. Initial testing on isolated errors refined the taxonomy, addressing challenges like overlapping categories. Final testing used "English as she is spoke" by Jose da Fonseca (1855), a text rich with authentic linguistic errors, to evaluate the system's capacity for handling complex, multi-layered analysis. The AI successfully identified diverse error types but showed limitations in contextual understanding and occasionally generated new error categories when encountering uncoded errors. This research demonstrates AI's potential to transform EFL instruction by automating detailed error analysis and feedback. While promising, further development is needed to improve contextual accuracy and expand the taxonomy to stylistic and discourse-level errors.
Read the full abstract

Computation and Language

Mitigating the Threshold Priming Effect in Large Language Model-Based Relevance Judgments via Personality Infusing

Nuo Chen, Hanpei Fang, Jiqun Liu et al. (+3 more)

Recent research has explored LLMs as scalable tools for relevance labeling, but studies indicate they are susceptible to priming effects, where prior relevance judgments influence later ones. Although psychological theories link personality traits to such biases, it is unclear whether simulated personalities in LLMs exhibit similar effects. We investigate how Big Five personality profiles in LLMs influence priming in relevance labeling, using multiple LLMs on TREC 2021 and 2022 Deep Learning Track datasets. Our results show that certain profiles, such as High Openness and Low Neuroticism, consistently reduce priming susceptibility. Additionally, the most effective personality in mitigating priming may vary across models and task types. Based on these findings, we propose personality prompting as a method to mitigate threshold priming, connecting psychological evidence with LLM-based evaluation practices.
Read the full abstract

All authors Nuo Chen, Hanpei Fang, Jiqun Liu, Wilson Wei, Tetsuya Sakai, Xiao-Ming Wu

Computation and Language · Cross-listed from cs.IR

Breaking It Down: Domain-Aware Semantic Segmentation for Retrieval Augmented Generation

Aparajitha Allamraju, Maitreya Prafulla Chitale, Hiranmai Sri Adibhatla et al. (+2 more)

Document chunking is a crucial component of Retrieval-Augmented Generation (RAG), as it directly affects the retrieval of relevant and precise context. Conventional fixed-length and recursive splitters often produce arbitrary, incoherent segments that fail to preserve semantic structure. Although semantic chunking has gained traction, its influence on generation quality remains underexplored. This paper introduces two efficient semantic chunking methods, Projected Similarity Chunking (PSC) and Metric Fusion Chunking (MFC), trained on PubMed data using three different embedding models. We further present an evaluation framework that measures the effect of chunking on both retrieval and generation by augmenting PubMedQA with full-text PubMed Central articles. Our results show substantial retrieval improvements (24x with PSC) in MRR and higher Hits@k on PubMedQA. We provide a comprehensive analysis, including statistical significance and response-time comparisons with common chunking libraries. Despite being trained on a single domain, PSC and MFC also generalize well, achieving strong out-of-domain generation performance across multiple datasets. Overall, our findings confirm that our semantic chunkers, especially PSC, consistently deliver superior performance.
Read the full abstract

All authors Aparajitha Allamraju, Maitreya Prafulla Chitale, Hiranmai Sri Adibhatla, Rahul Mishra, Manish Shrivastava

Computation and Language

CourseTimeQA: A Lecture-Video Benchmark and a Latency-Constrained Cross-Modal Fusion Method for Timestamped QA

Vsevolod Kovalev and Parteek Kumar

We study timestamped question answering over educational lecture videos under a single-GPU latency/memory budget. Given a natural-language query, the system retrieves relevant timestamped segments and synthesizes a grounded answer. We present CourseTimeQA (52.3 h, 902 queries across six courses) and a lightweight, latency-constrained cross-modal retriever (CrossFusion-RAG) that combines frozen encoders, a learned 512->768 vision projection, shallow query-agnostic cross-attention over ASR and frames with a temporal-consistency regularizer, and a small cross-attentive reranker. On CourseTimeQA, CrossFusion-RAG improves nDCG@10 by 0.10 and MRR by 0.08 over a strong BLIP-2 retriever while achieving approximately 1.55 s median end-to-end latency on a single A100. Closest comparators (zero-shot CLIP multi-frame pooling; CLIP + cross-encoder reranker + MMR; learned late-fusion gating; text-only hybrid with cross-encoder reranking and its MMR variant; caption-augmented text retrieval; non-learned temporal smoothing) are evaluated under matched hardware and indexing. We report robustness across ASR noise (WER quartiles), diagnostics for temporal localization, and full training/tuning details to support reproducible comparison.
Read the full abstract

Computation and Language

IndicParam: Benchmark to evaluate LLMs on low-resource Indic Languages

Ayush Maheshwari, Kaushal Sharma, Vivek Patel et al. (+1 more)

While large language models excel on high-resource multilingual tasks, low- and extremely low-resource Indic languages remain severely under-evaluated. We present IndicParam, a human-curated benchmark of over 13,000 multiple-choice questions covering 11 such languages (Nepali, Gujarati, Marathi, Odia as low-resource; Dogri, Maithili, Rajasthani, Sanskrit, Bodo, Santali, Konkani as extremely low-resource) plus Sanskrit-English code-mixed set. We evaluated 20 LLMs, both proprietary and open-weights, which reveals that even the top-performing Gemini-2.5 reaches 58% average accuracy, followed by GPT-5 (45) and DeepSeek-3.2 (43.1). We additionally label each question as knowledge-oriented or purely linguistic to discriminate factual recall from grammatical proficiency. Further, we assess the ability of LLMs to handle diverse question formats-such as list-based matching, assertion-reason pairs, and sequence ordering-alongside conventional multiple-choice questions. IndicParam provides insights into limitations of cross-lingual transfer and establishes a challenging benchmark for Indic languages. The dataset is available at https://huggingface.co/datasets/bharatgenai/IndicParam. Scripts to run benchmark are present at https://github.com/ayushbits/IndicParam.
Read the full abstract

All authors Ayush Maheshwari, Kaushal Sharma, Vivek Patel, Aditya Maheshwari

Computation and Language

Assertion-Conditioned Compliance: A Provenance-Aware Vulnerability in Multi-Turn Tool-Calling Agents

Daud Waqas, Aaryamaan Golthi, Erika Hayashida et al. (+1 more)

Multi-turn tool-calling LLMs (models capable of invoking external APIs or tools across several user turns) have emerged as a key feature in modern AI assistants, enabling extended dialogues from benign tasks to critical business, medical, and financial operations. Yet implementing multi-turn pipelines remains difficult for many safety-critical industries due to ongoing concerns regarding model resilience. While standardized benchmarks such as the Berkeley Function-Calling Leaderboard (BFCL) have underpinned confidence concerning advanced function-calling models (like Salesforce's xLAM V2), there is still a lack of visibility into multi-turn conversation-level robustness, especially given their exposure to real-world systems. In this paper, we introduce Assertion-Conditioned Compliance (A-CC), a novel evaluation paradigm for multi-turn function-calling dialogues. A-CC provides holistic metrics that evaluate a model's behavior when confronted with misleading assertions originating from two distinct vectors: (1) user-sourced assertions (USAs), which measure sycophancy toward plausible but misinformed user beliefs, and (2) function-sourced assertions (FSAs), which measure compliance with plausible but contradictory system policies (e.g., stale hints from unmaintained tools). Our results show that models are highly vulnerable to both USA sycophancy and FSA policy conflicts, confirming A-CC as a critical, latent vulnerability in deployed agents.
Read the full abstract

All authors Daud Waqas, Aaryamaan Golthi, Erika Hayashida, Huanzhi Mao

Computation and Language

Evidence-Guided Schema Normalization for Temporal Tabular Reasoning

Ashish Thanga, Vibhu Dixit, Abhilash Shankarampeta et al. (+1 more)

Temporal reasoning over evolving semi-structured tables poses a challenge to current QA systems. We propose an approach that recasts the task as automated knowledge base construction: (1) prompting an LLM to synthesize a 3NF-compliant relational schema from Wikipedia infobox timelines, (2) populating the schema to obtain a queryable database, and (3) generating and executing SQL queries against it, with QA accuracy serving as an extrinsic evaluation of the constructed knowledge base. In a controlled grid of three schema generators crossed with six query models, the schema source accounts for 79.5% of the exact match (EM) variance against 1.6% for the query model: replacing the schema, and the prompt scaffolding derived from it, shifts EM by 14.7 to 20.0 points, whereas replacing the query model under a fixed schema shifts it by 4.4 to 12.1. From this evidence, we distill three candidate schema-design principles: balanced normalization, semantic naming, and consistent temporal anchoring, framed as correlational hypotheses. Our best configuration (Gemini 2.5 Flash schemas + Gemini-2.0-Flash queries) reaches 80.39 EM, 11.5 points above the strongest reported baseline (68.89 EM); an open-weights configuration reaches 79.52.
Read the full abstract

All authors Ashish Thanga, Vibhu Dixit, Abhilash Shankarampeta, Vivek Gupta