Search: “arxiv-cs-cl”

On the Generalization Gap in Self-Evolving Language Model Reasoning

arXiv:2606.01075v1 Announce Type: new Abstract: Recent work suggests that large language models (LLMs) can improve through self-evolution (SE), using supervision signals generated by the model itself.

Research

On the Salience of Low-Probability Tokens for AI-Generated Text Detection: A Multiscale Uncertainty Perspective

DGX agent

arXiv:2606.02158v1 Announce Type: new Abstract: AI-generated text increasingly blends with human writing, raising practical risks such as misinformation, academic misuse, and corpora contamination. Wh

researcharxiv-cs-cl

2 Jun 2026

On the Scaling of PEFT: Towards Million Personal Models of Trillion Parameters

arXiv:2606.02437v1 Announce Type: cross Abstract: Parameter-efficient fine-tuning (PEFT) is usually treated as a cheaper alternative to full fine-tuning. We study a broader role: small trainable adapt

OncoReason: Structuring Clinical Reasoning in LLMs for Robust and Interpretable Survival Prediction

arXiv:2510.17532v2 Announce Type: replace Abstract: Predicting cancer treatment outcomes requires models that are both accurate and interpretable, particularly in the presence of heterogeneous clinica

PaperVoyager : Building Interactive Web with Visual Language Models

arXiv:2603.22999v3 Announce Type: replace Abstract: Recent advances in visual language models have enabled autonomous agents for complex reasoning, tool use, and document understanding. However, exist

Parameter Alignment Mitigates Catastrophic Forgetting in Multilingual Expert Language Models

arXiv:2606.00284v1 Announce Type: new Abstract: While continual pretraining~(CPT) is a practical way to extend large language models to new languages, naive finetuning on targeted data erodes existing

Applications

Parametric Social Identity Injection and Diversification in Public Opinion Simulation

arXiv:2603.16142v2 Announce Type: replace Abstract: Large language models (LLMs) have recently been adopted as synthetic agents for public opinion simulation, offering a promising alternative to costl

applicationsarxiv-cs-cl

Peacemaker at ATE-IT: Automatic term extraction from Italian text for waste management data using encoder model

arXiv:2606.01469v1 Announce Type: new Abstract: The development of automatic term extraction has become increasingly important in modern technology. Automatic term extraction can be found in virtually

Phoneme-Level Visual Speech Recognition via Point-Visual Fusion and Language Model Reconstruction

arXiv:2507.18863v2 Announce Type: replace-cross Abstract: Visual Automatic Speech Recognition (V-ASR) is a challenging task that involves interpreting spoken language solely from visual information, s

PMC-InterCPT: Rethinking Biomedical Interleaved Data for Multimodal Continued Pretraining

arXiv:2606.01049v1 Announce Type: new Abstract: Large-scale biomedical image-text datasets extracted from scientific literature provide valuable resources for medical multimodal model training. These

Hardware

PortBERT: Navigating the Depths of Portuguese Language Models

arXiv:2606.02100v1 Announce Type: new Abstract: Transformer models dominate modern NLP, but efficient, language-specific models remain scarce. In Portuguese, most focus on scale or accuracy, often neg

hardwarearxiv-cs-cl

Pramana: Fine-Tuning Large Language Models for Epistemic Reasoning through Navya-Nyaya

arXiv:2604.04937v1 Announce Type: cross Abstract: Large language models produce fluent text but struggle with systematic reasoning, often hallucinating confident but unfounded claims. When Apple resea

Local Ai

ProactiveLLM: Learning Active Interaction for Streaming Large Language Models

arXiv:2606.00523v1 Announce Type: new Abstract: Standard Large Language Models (LLMs) follow a read-then-generate paradigm, causing unnecessary latency and computation. Streaming LLMs alleviate this i

local-aiarxiv-cs-cl

ProtStructQA: A Denotation Threshold in Protein Structural Reasoning

arXiv:2606.00451v1 Announce Type: new Abstract: Protein-language systems are often evaluated by whether they generate plausible biological text, but a structural question has a sharper semantics: it d

Quality-Diversity Evolution for Discovering Diverse Vulnerabilities in LLM Safety

arXiv:2606.00801v1 Announce Type: cross Abstract: Current approaches to LLM adversarial testing suffer from coverage gaps: manual red-teaming does not scale, LLM-as-attacker methods exhibit mode colla

R2-Router: A New Paradigm for LLM Routing with Reasoning

arXiv:2602.02823v2 Announce Type: replace Abstract: As LLMs proliferate with diverse capabilities and costs, LLM routing has emerged by learning to predict each LLM's quality and cost for a given quer

Tutorials

RCEM: Embedder Equipped with Query Rewriting Skill for Robust Conversational Search in Distributional Shift

arXiv:2606.01697v1 Announce Type: new Abstract: Conversational search has become increasingly important in retrieval-augmented generation (RAG) systems, where users interact with AI assistants through

tutorialsarxiv-cs-cl

RealityTest: How People Probe AI Identity and Whether Models Disclose It

arXiv:2606.00168v1 Announce Type: new Abstract: AI systems are increasingly deployed in conversational settings where users may be uncertain whether they are speaking with a human or an AI. Despite mo

Reasmory: 3D Reconstruction as Explicit Memory for VLMs Spatial Reasoning

arXiv:2606.00963v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) exhibit emerging spatial reasoning capabilities, yet they remain unreliable on tasks requiring precise spatial understan

Reconsidering Positional Supervision in Masked Diffusion Language Model Training

arXiv:2601.22947v2 Announce Type: replace Abstract: Masked diffusion language models (MDLMs) generate text by unmasking tokens in parallel and have recently emerged as alternatives to autoregressive l

RedDebate: Safer Responses Through Multi-Agent Red Teaming Debates

arXiv:2506.11083v3 Announce Type: replace Abstract: We introduce RedDebate, a novel multi-agent debate framework that provides the foundation for Large Language Models (LLMs) to identify and mitigate

RenoBench: A Citation Parsing Benchmark

arXiv:2603.25640v2 Announce Type: replace-cross Abstract: Accurate parsing of citations is necessary for machine-readable scholarly infrastructure. But, despite sustained interest in this problem, exi

ResMerge: Residual-based Spectral Merging of Large Language Models

arXiv:2606.02252v1 Announce Type: new Abstract: Model merging offers a training-free way to combine multiple post-trained expert models, but merging experts obtained through reinforcement learning (RL

Resonant Context Anchoring: Decoupling Attention Routing and Signal Gain at Inference Time

arXiv:2606.01923v1 Announce Type: new Abstract: Large Language Models (LLMs) frequently exhibit 'contextual disregard' when faced with input evidence that conflicts with their internal parametric memo

Revise, Don't Freeze: Sampler-Matched Training for Self-Correcting Masked Diffusion Language Models

arXiv:2606.01026v1 Announce Type: new Abstract: Masked diffusion language models (MDLMs) re-predict every position at each denoising step, but standard samplers commit tokens once revealed, leaving th

RoboTrustBench: Benchmarking the Trustworthiness of Video World Models for Robotic Manipulation

arXiv:2606.01600v1 Announce Type: cross Abstract: Video world models are increasingly used in robotic manipulation, yet existing benchmarks mostly evaluate them under valid, feasible, and safe instruc

Applications

Robust Asynchronous Planning via Auto-Formalization

arXiv:2606.00981v1 Announce Type: new Abstract: LLMs can plan by either generating action sequences directly as a Planner or translating tasks into domain specific language for an external solver as a

applicationsarxiv-cs-cl

Tutorials

Robust Reasoning via Dynamic Token Selection for Distribution-Aligned Self-Distillation

arXiv:2606.00628v1 Announce Type: new Abstract: Self-distillation improves learning efficiency by rewriting reference answers as training data that better matches the model's own distribution. However

tutorialsarxiv-cs-cl

SALSA: Speech Aware LLM Adaptation via Learned Steering Activation Vectors

arXiv:2606.00460v1 Announce Type: new Abstract: Speech-aware large language models often generalize poorly to out-of-domain settings. We propose SALSA (Speech-Aware LLM Adaptation via Learned Steering

Same Payload, Different Channel: Measuring Trust Asymmetry in Tool-Using Language Models

arXiv:2606.00566v1 Announce Type: cross Abstract: As language models take on agentic roles that span calling external APIs, reading tool outputs, and acting on instructions embedded in third-party con

Sandboxed Coding Agents are Competitive Omni-modal Task Solvers

arXiv:2606.00579v1 Announce Type: new Abstract: As multimodal LLMs increasingly target video and audio, it is often assumed that such tasks require native omnimodal models. We show that this is not al

SARA: Stress Test Reasoning in Audio Deepfake Detection

arXiv:2601.03615v2 Announce Type: replace Abstract: Audio Language Models (ALMs) offer a promising shift towards explainable audio deepfake detections (ADD), moving beyond extit{black-box} classifiers

Agents

Scaling Agentic Capabilities via Grounded Interaction Synthesis

arXiv:2606.02001v1 Announce Type: new Abstract: General agentic intelligence hinges on the ability to interact with diverse real-world tools to complete complex tasks, a capability fundamentally tied

agentsarxiv-cs-cl

Agents

SciAgentGym: Benchmarking Multi-Step Scientific Tool-use in LLM Agents

arXiv:2602.12984v2 Announce Type: replace Abstract: Scientific reasoning inherently demands integrating sophisticated toolkits to navigate domain-specific knowledge. Yet, current benchmarks largely ov

agentsarxiv-cs-cl

Search-on-Graph: Iterative Informed Navigation for Large Language Model Reasoning on Knowledge Graphs

arXiv:2510.08825v2 Announce Type: replace Abstract: Large language models (LLMs) augmented with knowledge graphs (KGs) offer a promising approach for knowledge-intensive reasoning. Central to this app

Seeing Through the MiRAGE: Evaluating Multimodal Retrieval Augmented Generation

arXiv:2510.24870v2 Announce Type: replace Abstract: We introduce MiRAGE, an evaluation framework for retrieval-augmented generation (RAG) from multimodal sources. As audiovisual media becomes a preval

SentGuard: Sentence-Level Streaming Guardrails for Large Language Models

arXiv:2606.02041v1 Announce Type: new Abstract: Large language models increasingly stream long, reasoning-intensive responses in real time, making when to moderate as critical as whether to moderate.

SindBERT, the Sailor: Charting the Seas of Turkish NLP

arXiv:2510.21364v2 Announce Type: replace Abstract: Transformer models have revolutionized NLP, yet many morphologically rich languages remain underrepresented in large-scale pre-training efforts. Wit

SkillHarm: Lifecycle-Aware Skill-Based Attacks via Automated Construction

arXiv:2606.02540v1 Announce Type: new Abstract: Agent skills occupy a privileged position in the agent workflow, as agents are expected to implicitly follow and execute them, rendering third-party ski

SmartThinker: Progressive Chain-of-Thought Length Calibration for Efficient Large Language Model Reasoning

arXiv:2603.08000v2 Announce Type: replace Abstract: Large reasoning models (LRMs) like OpenAI o1 and DeepSeek-R1 achieve high accuracy on complex tasks by adopting long chain-of-thought (CoT) reasonin

SN-WER: Script-Normalized WER for Multi-Script Indic ASR Evaluation

arXiv:2606.02548v1 Announce Type: new Abstract: Word Error Rate (WER) is the dominant metric for automatic speech recognition (ASR), but it can overestimate errors when references and hypotheses encod

Sparse Autoencoders for Interpretable Emotion Control in Text-to-Speech

arXiv:2606.01479v1 Announce Type: new Abstract: Integrating large language models (LLMs) into text-to-speech (TTS) systems has improved speech expressiveness, yet interpretable emotional control remai

Stabilizing Policy Optimization via Logits Convexity

arXiv:2603.00963v2 Announce Type: replace-cross Abstract: While reinforcement learning (RL) has been central to the recent success of large language models (LLMs), RL optimization is notoriously unsta

Structured Semantic Information Helps Retrieve Better Examples for In-Context Learning Applied to Few-Shot Relation Extraction

arXiv:2601.20803v2 Announce Type: replace Abstract: This paper presents several strategies to automatically obtain additional examples for in-context learning, effectively transforming relation extrac

Agents

SWE-rebench V2: Language-Agnostic SWE Task Collection at Scale

arXiv:2602.23866v2 Announce Type: replace-cross Abstract: Software engineering agents (SWE) are improving rapidly, with recent gains largely driven by reinforcement learning (RL). However, RL training

agentsarxiv-cs-cl

Sympatheia: Emotionally Adaptive Voice Assistant with Continuous Affect Conditioning

arXiv:2606.00851v1 Announce Type: cross Abstract: Empathetic spoken dialogue systems must infer a user's emotional state to respond appropriately, yet everyday speech often carries weak, neutral, or a

TalkTag: Fine-Grained Morphosyntactic Error Annotation for Transcribed Speech

arXiv:2606.01820v1 Announce Type: new Abstract: Fine-grained morphosyntactic error annotation is important in clinical and developmental language research, yet it is labour-intensive, expert-dependent

Task Structure Reverses Layerwise State Encoding in Sequence Models

arXiv:2606.00926v1 Announce Type: cross Abstract: Mechanistic studies of sequence models often treat layerwise state encodings as architectural traits: recurrent models concentrate readable state, att