Search: “arxiv-cs-cl”

NodeSynth: Socially Aligned Synthetic Data for AI Evaluation

arXiv:2605.14381v1 Announce Type: cross Abstract: Recent advancements in generative AI facilitate large-scale synthetic data generation for model evaluation. However, without targeted approaches, thes

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Pause and Reflect: Conformal Aggregation for Chain-of-Thought Reasoning

DGX agent

arXiv:2605.14098v1 Announce Type: cross Abstract: Chain-of-thought (CoT) reasoning with self-consistency improves performance by aggregating multiple sampled reasoning paths. In this setting, correctn

researcharxiv-cs-cl

15 May 2026

Performance-Driven Policy Optimization for Speculative Decoding with Adaptive Windowing

arXiv:2605.14978v1 Announce Type: new Abstract: Speculative decoding accelerates LLM inference by having a lightweight draft model propose speculative windows of candidate tokens for parallel verifica

Persian MusicGen: A Large-Scale Dataset and Culturally-Aware Generative Model for Persian Music

arXiv:2605.14765v1 Announce Type: cross Abstract: Persian music, with its unique tonalities, modal systems (Dastgah), and rhythmic structures, presents significant challenges for music generation mode

Physics-R1: An Audited Olympiad Corpus and Recipe for Visual Physics Reasoning

arXiv:2605.14040v1 Announce Type: new Abstract: We audit the multimodal-physics evaluation pipeline end-to-end and document three undetected construction practices that distort how the field measures

Tutorials

Polar probe linearly decodes semantic structures from LLMs

arXiv:2605.14125v1 Announce Type: new Abstract: How do artificial neural networks bind concepts to form complex semantic structures? Here, we propose a simple neural code, whereby the existence and th

tutorialsarxiv-cs-cl

Proposal and study of statistical features for string similarity computation and classification

arXiv:2605.15110v1 Announce Type: cross Abstract: Adaptations of features commonly applied in the field of visual computing, co-occurrence matrix (COM) and run-length matrix (RLM), are proposed for th

Proxy Compression for Language Modeling

arXiv:2602.04289v2 Announce Type: replace Abstract: Modern language models are trained almost exclusively on token sequences produced by a fixed tokenizer, an external lossless compressor often over U

QOuLiPo: What a quantum computer sees when it reads a book

arXiv:2605.14188v1 Announce Type: cross Abstract: What does a book look like to a quantum computer? This paper takes eight classical works of the Renaissance and its late-antique inheritance -- from A

Reasoning Model Is Superior LLM-Judge, Yet Suffers from Biases

arXiv:2601.03630v2 Announce Type: replace Abstract: This paper presents the first systematic comparison investigating whether Large Reasoning Models (LRMs) are superior judges to non-reasoning LLMs. O

Reinforcement Learning with Semantic Rewards Enables Low-Resource Language Expansion without Alignment Tax

arXiv:2605.14366v1 Announce Type: new Abstract: Extending large language models (LLMs) to low-resource languages often incurs an 'alignment tax': improvements in the target language come at the cost o

Agents

Remember Your Trace: Memory-Guided Long-Horizon Agentic Framework for Consistent and Hierarchical Repository-Level Code Documentation

arXiv:2605.14563v1 Announce Type: cross Abstract: Automated code documentation is essential for modern software development, providing the contextual grounding that both human developers and coding ag

agentsarxiv-cs-cl

Rethinking Layer Relevance in Large Language Models Beyond Cosine Similarity

arXiv:2605.14075v1 Announce Type: cross Abstract: Large language models (LLMs) have revolutionized natural language processing. Understanding their internal mechanisms is crucial for developing more i

SciPaths: Forecasting Pathways to Scientific Discovery

arXiv:2605.14600v1 Announce Type: new Abstract: Scientific progress depends on sequences of enabling contributions, yet existing AI4Science benchmarks largely focus on citation prediction, literature

Local Ai

Self-Pruned Key-Value Attention: Learning When to Write by Predicting Future Utility

arXiv:2605.14037v1 Announce Type: cross Abstract: Under modern test-time compute and agentic paradigms, language models process ever-longer sequences. Efficient text generation with transformer archit

local-aiarxiv-cs-cl

Small Language Models (SLMs) Can Still Pack a Punch: A survey (updated 2026)

arXiv:2501.05465v2 Announce Type: replace Abstract: As foundation AI models continue to increase in size, an important question arises - is massive scale the only path forward? This survey of about 16

Talk is (Not) Cheap: A Taxonomy and Benchmark Coverage Audit for LLM Attacks

arXiv:2605.15118v1 Announce Type: cross Abstract: We introduce a reusable framework for auditing whether LLM attack benchmarks collectively cover the threat surface: a 4imes6 Target imes Technique mat

The Scientific Contribution Graph: Automated Literature-based Technological Roadmapping at Scale

arXiv:2605.15011v1 Announce Type: new Abstract: Scientific contributions rarely develop in isolation, but instead build upon prior discoveries. We formulate the task of automated technological roadmap

Think When Needed: Adaptive Reasoning-Driven Multimodal Embeddings with a Dual-LoRA Architecture

arXiv:2605.14448v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) have emerged as a powerful backbone for multimodal embeddings. Recent methods introduce chain-of-thought (CoT

Tokenizer Fertility and Zero-Shot Performance of Foundation Models on Ukrainian Legal Text: A Comparative Study

arXiv:2605.14890v1 Announce Type: new Abstract: Foundation models tokenize Ukrainian legal text with vastly different efficiency, yet no systematic comparison exists for this domain. We benchmark seve

TRIM: Token-wise Attention-Derived Saliency for Data-Efficient Instruction Tuning

arXiv:2510.07118v3 Announce Type: replace Abstract: Instruction tuning is essential for aligning large language models (LLMs) to downstream tasks and commonly relies on large, diverse corpora. However

Uncertainty Quantification for Large Language Diffusion Models

arXiv:2605.14570v1 Announce Type: new Abstract: Large Language Diffusion Models (LLDMs) are emerging as an alternative to autoregressive models, offering faster inference through higher parallelism. S

VectraYX-Nano: A 42M-Parameter Spanish Cybersecurity Language Model with Curriculum Learning and Native Tool Use

arXiv:2605.13989v1 Announce Type: new Abstract: We present VectraYX-Nano, a 41.95M-parameter decoder-only language model trained from scratch in Spanish for cybersecurity, with a Latin-American focus

What Do AI Agents Talk About? Discourse and Architectural Constraints in the First AI-Only Social Network

arXiv:2603.07880v5 Announce Type: replace Abstract: Moltbook is the first large-scale social network built for autonomous AI agent-to-agent interaction. Early studies on Moltbook have interpreted its

Applications

What Makes Words Hard? Sakura at BEA 2026 Shared Task on Vocabulary Difficulty Prediction

arXiv:2605.14257v1 Announce Type: new Abstract: We describe two types of models for vocabulary difficulty prediction: a high-accuracy black-box model, which achieved the top shared task result in the

applicationsarxiv-cs-cl

When Evidence Conflicts: Uncertainty and Order Effects in Retrieval-Augmented Biomedical Question Answering

arXiv:2605.14115v1 Announce Type: new Abstract: Biomedical retrieval-augmented large language models (LLMs) often face evidence that is incomplete, misleading, or internally contradictory, yet evaluat

A categorical error sensitivity index (ISEC): A preventive ordinal decision-support measure for irrecoverable errors in manual data entry systems

arXiv:2605.12328v1 Announce Type: new Abstract: Data entry systems remain structurally vulnerable to categorical misclassifications, particularly in small and medium sized enterprises (SMEs). When nom

A Causal Language Modeling Detour Improves Encoder Continued Pretraining

arXiv:2605.12438v1 Announce Type: new Abstract: When adapting an encoder to a new domain, the standard approach is to continue training with Masked Language Modeling (MLM). We show that temporarily sw

A Comparative Study of Controlled Text Generation Systems Using Level-Playing-Field Evaluation Principles

arXiv:2605.12395v1 Announce Type: new Abstract: Background: Many different approaches to controlled text generation (CTG) have been proposed over recent years, but it is difficult to get a clear pictu

A Formal Comparison Between Chain of Thought and Latent Thought

arXiv:2509.25239v3 Announce Type: replace-cross Abstract: Chain of thought (CoT) elicits reasoning in large language models by explicitly generating intermediate tokens. In contrast, latent thought re

A Study on Hidden Layer Distillation for Large Language Model Pre-Training

arXiv:2605.11513v1 Announce Type: new Abstract: Knowledge Distillation (KD) is a critical tool for training Large Language Models (LLMs), yet the majority of research focuses on approaches that rely s

A Survey of On-Policy Distillation for Large Language Models

arXiv:2604.00626v2 Announce Type: replace-cross Abstract: As Large Language Models (LLMs) continue to grow in both capability and cost, transferring frontier capabilities into smaller, deployable stud

A Theoretical Analysis of Why Masked Diffusion Models Mitigate the Reversal Curse

arXiv:2602.02133v2 Announce Type: replace-cross Abstract: Autoregressive language models (ARMs) suffer from the reversal curse: after learning ''A is B,'' they often fail on the reverse query ''B is A

A Theory of Time-Sensitive Language Generation: Sparse Hallucination Beats Mode Collapse

arXiv:2605.11302v1 Announce Type: cross Abstract: We study language generation in the limit under a global preference ordering on strings, as introduced by Kleinberg and Wei. As in [arXiv:2504.14370,

AcuityBench: Evaluating Clinical Acuity Identification and Uncertainty Alignment

arXiv:2605.11398v1 Announce Type: cross Abstract: We introduce AcuityBench, a benchmark for evaluating whether language models identify the appropriate urgency of care from user medical presentations.

Ada-MK: Adaptive MegaKernel Optimization via Automated DAG-based Search for LLM Inference

arXiv:2605.11581v1 Announce Type: new Abstract: When large language models (LLMs) serve real-time inference in commercial online advertising systems, end-to-end latency must be strictly bounded to the

Adaptive Contrastive Learning on Multimodal Transformer for Review Helpfulness Predictions

arXiv:2211.03524v2 Announce Type: replace Abstract: Modern Review Helpfulness Prediction systems are dependent upon multiple modalities, typically texts and images. Unfortunately, those contemporary a

Adaptive Teacher Exposure for Self-Distillation in LLM Reasoning

arXiv:2605.11458v1 Announce Type: cross Abstract: On-policy self-distillation has become a strong recipe for LLM reasoning, where a privileged teacher supervises the student's own rollouts while condi

Agent-BRACE: Decoupling Beliefs from Actions in Long-Horizon Tasks via Verbalized State Uncertainty

arXiv:2605.11436v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed on long-horizon tasks in partially observable environments, where they must act while inferring a

AgentDisCo: Towards Disentanglement and Collaboration in Open-ended Deep Research Agents

arXiv:2605.11732v1 Announce Type: cross Abstract: In this paper, we present AgentDisCo, a novel Disentangled and Collaborative agentic architecture that formulates deep research as an adversarial opti

Agents

AgentShield: Deception-based Compromise Detection for Tool-using LLM Agents

arXiv:2605.11026v1 Announce Type: cross Abstract: Defenses against indirect prompt injection (IPI) in tool-using LLM agents share two structural weaknesses. First, they all attempt to prevent attacks

agentsarxiv-cs-cl

Allegory of the Cave: Measurement-Grounded Vision-Language Learning

arXiv:2605.11727v1 Announce Type: cross Abstract: Vision-language models typically reason over post-ISP RGB images, although RGB rendering can clip, suppress, or quantize sensor evidence before infere

An Empirical Study of Automating Agent Evaluation

arXiv:2605.11378v1 Announce Type: new Abstract: Agent evaluation requires assessing complex multi-step behaviors involving tool use and intermediate reasoning, making it costly and expertise-intensive

Anti-Self-Distillation for Reasoning RL via Pointwise Mutual Information

arXiv:2605.11609v1 Announce Type: cross Abstract: On-policy self-distillation, where a student is pulled toward a copy of itself conditioned on privileged context (e.g., a verified solution or feedbac

Asymmetric Advantage Modulation Calibrates Entropy Dynamics in RLVR

arXiv:2604.04894v2 Announce Type: replace Abstract: Reinforcement learning with verifiable rewards (RLVR) has substantially improved the reasoning ability of large language models (LLMs), but it often

Hardware

AutoLLMResearch: Training Research Agents for Automating LLM Experiment Configuration -- Learning from Cheap, Optimizing Expensive

arXiv:2605.11518v1 Announce Type: cross Abstract: Effectively configuring scalable large language model (LLM) experiments, spanning architecture design, hyperparameter tuning, and beyond, is crucial f

hardwarearxiv-cs-cl

AutoMonitor-Bench: Evaluating the Reliability of LLM-Based Misbehavior Monitor

arXiv:2601.05752v3 Announce Type: replace Abstract: We introduce AutoMonitor-Bench, the first benchmark designed to systematically evaluate the reliability of LLM-based misbehavior monitors across div

BEExformer: A Fast Inferencing Binarized Transformer with Early Exits

arXiv:2412.05225v3 Announce Type: replace Abstract: Large Language Models (LLMs) based on transformers achieve cutting-edge results on a variety of applications. However, their enormous size and proce