{"total":259,"rows":[{"id":"2608.30320","title":"On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability","created":"2026-08-31","cited_by_count":35,"area":"moe"},{"id":"2607.15161","title":"On-Policy Delta Distillation","created":"2026-07-16","cited_by_count":15,"area":"alignment"},{"id":"2607.04884","title":"HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Better","created":"2026-07-06","cited_by_count":14,"area":"mllm"},{"id":"2608.27454","title":"WikiSkill: Compiling Agent Experience into Persistent Knowledge for Skill Evolution","created":"2026-08-27","cited_by_count":12,"area":null},{"id":"2607.22083","title":"Nanbeige4.2-3B: Unlocking Agentic Capabilities in a Compact Model","created":"2026-07-24","cited_by_count":11,"area":"agents"},{"id":"2608.00146","title":"DiffusionGemma Technical Report","created":"2026-07-31","cited_by_count":11,"area":"diffusion"},{"id":"2608.09096","title":"Evo-Bench: Can Language Models Improve Agent Harness?","created":"2026-08-10","cited_by_count":11,"area":"agents"},{"id":"2608.29910","title":"Matrix-Game 3.5: Enhancing Real-Time Streaming Interactive World Models with Patch Memory","created":"2026-08-30","cited_by_count":9,"area":"wm"},{"id":"2609.00111","title":"Qwen-Drive-1.0: An Initial Step towards a Vision-Language Foundation Model for Autonomous Driving","created":"2026-08-31","cited_by_count":9,"area":"mllm"},{"id":"2607.26760","title":"Metis: Memory Foundation Model","created":"2026-07-29","cited_by_count":7,"area":"lr"},{"id":"2609.11929","title":"SenseNova-U1.5: Towards Native Unified Visual Intelligence","created":"2026-09-10","cited_by_count":7,"area":"distill"},{"id":"2607.19829","title":"DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection","created":"2026-07-22","cited_by_count":6,"area":"safety"},{"id":"2608.26105","title":"VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning","created":"2026-08-26","cited_by_count":6,"area":"lr"},{"id":"2608.15045","title":"MOSS-VL Technical Report","created":"2026-08-15","cited_by_count":5,"area":"mllm"},{"id":"2609.04565","title":"Extremely Sparse Supervision Incentivizes Reasoning Ability","created":"2026-09-03","cited_by_count":5,"area":"lr"},{"id":"2609.30243","title":"JevOut: Natural Context Can Flip Decision Models","created":"2026-09-24","cited_by_count":4,"area":null},{"id":"2608.02287","title":"SKT: Skill-Use Training at Scale via Verified Synthetic Data Generation","created":"2026-08-03","cited_by_count":3,"area":"agents"},{"id":"2608.10905","title":"ReOrder-OPD:Reliability-Aware Prompt Ordering for On-Policy Distillation","created":"2026-08-11","cited_by_count":3,"area":"distill"},{"id":"2608.19147","title":"Pre-Compiled Pipeline Shards for Distributed LLM Inference on Intel AI PC Fleets","created":"2026-08-19","cited_by_count":3,"area":"inference"},{"id":"2607.23806","title":"A Frozen 12B Beats Frontier Models on Verified Work: 100% Accuracy, 0 Tokens, Bit-Exact, Forever","created":"2026-07-26","cited_by_count":2,"area":null},{"id":"2608.16211","title":"BaT: Towards Self-Evolving Medical Research Agent with Stage Rubrics","created":"2026-08-17","cited_by_count":2,"area":"agents"},{"id":"2609.04290","title":"Evidence Integration in Large Language Models","created":"2026-09-03","cited_by_count":2,"area":"rag"},{"id":"2609.06746","title":"Reason Through the Latent! Making Latent Visual Reasoning Necessary","created":"2026-09-06","cited_by_count":2,"area":"lr"},{"id":"2609.35621","title":"Cartridges++: KV Cache Compression without Off-Context Derailment","created":"2026-09-28","cited_by_count":2,"area":"inference"},{"id":"2607.20327","title":"PyroDash: Cost-Efficient Token-Level Small-Large Language Model Collaborative Inference","created":"2026-07-22","cited_by_count":1,"area":"lr"},{"id":"2607.28175","title":"AgenticASR: Refining Speech Recognition in Real-World Scenarios via an Agentic Approach","created":"2026-07-30","cited_by_count":1,"area":"agents"},{"id":"2607.28707","title":"Demystifying Entropy-based Selection for Chain-of-Thought Compression in Large Reasoning Models","created":"2026-07-30","cited_by_count":1,"area":"lr"},{"id":"2608.00200","title":"TRACE-TS: Attribution-Grounded and Traceable Sensor-Language Reasoning for Human Activity Understanding","created":"2026-07-31","cited_by_count":1,"area":"lr"},{"id":"2608.10108","title":"MESA:Task-Adaptive Multi-Structure Evidence Selection for Long-Horizon Agent Memory","created":"2026-08-10","cited_by_count":1,"area":null},{"id":"2608.10908","title":"Order Matters: LVLMs as Judges for Temporal Reasoning in Image Sequences","created":"2026-08-11","cited_by_count":1,"area":"lr"},{"id":"2608.12547","title":"Do LLMs Beat Nash? Testing Decentralized Coordination in Self-Play Multi-Agent Games","created":"2026-08-12","cited_by_count":1,"area":"agents"},{"id":"2608.13482","title":"Synthetic Persona Pretraining: Alignment from Token Zero","created":"2026-08-13","cited_by_count":1,"area":null},{"id":"2608.13966","title":"QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction","created":"2026-08-14","cited_by_count":1,"area":null},{"id":"2608.17270","title":"Do LLMs Know a Good Hypothesis When They See One? Logit-Based Energy Scoring Outperforms Prompted LLM-as-Judge for Scientific Hypothesis Ranking","created":"2026-08-18","cited_by_count":1,"area":null},{"id":"2608.20631","title":"Weighted Memory Tree: Remembering What Matters for Long-Horizon LLM Agents","created":"2026-08-21","cited_by_count":1,"area":"agents"},{"id":"2608.24231","title":"RecurSE: Bounded Recursive Self-Evaluation for LLM Rubric Judges","created":"2026-08-25","cited_by_count":1,"area":"alignment"},{"id":"2608.26582","title":"J-Zero: Unified Challenger--Solver--Judge Self-Evolution from Zero Data","created":"2026-08-27","cited_by_count":1,"area":null},{"id":"2608.27370","title":"Puro-2B: Poor Lab's Qwen2-1.5B Trained on RTX 5090 within $5090","created":"2026-08-27","cited_by_count":1,"area":null},{"id":"2608.30413","title":"DERELAB: Probing Defeasible Reasoning and Confirmation Bias in LLMs with a Generative Benchmark","created":"2026-08-31","cited_by_count":1,"area":"lr"},{"id":"2609.00251","title":"Hypotheses-Guided Self Distillation for Continual Personalization","created":"2026-08-31","cited_by_count":1,"area":"distill"},{"id":"2609.02737","title":"Language Models Can Control Their Own Attention","created":"2026-09-02","cited_by_count":1,"area":"inference"},{"id":"2609.11875","title":"UniMPA: A Unified Memory-Prediction-Action Model via Action-Grounded Transition Modeling","created":"2026-09-10","cited_by_count":1,"area":"vla"},{"id":"2609.14717","title":"Carryover Drafting: Recycling Rejected States for Speculative Decoding","created":"2026-09-13","cited_by_count":1,"area":"inference"},{"id":"2609.15982","title":"The Router Within: Eliciting Native Skill Routing from a Frozen LLM","created":"2026-09-14","cited_by_count":1,"area":"agents"},{"id":"2609.17496","title":"Verifiable Social Reasoning for LLM Assistants","created":"2026-09-15","cited_by_count":1,"area":"lr"},{"id":"2609.19745","title":"Region-Level Policy Optimization for Fine-grained MLLM Perception","created":"2026-09-17","cited_by_count":1,"area":null},{"id":"2609.19991","title":"AVTrace: Diagnosing Audio-Visual Temporal Reasoning in Omni Models","created":"2026-09-17","cited_by_count":1,"area":"lr"},{"id":"2609.20648","title":"SkipVLA: Skipping VLA Steps with Classical Planning for Fast Robot Manipulation","created":"2026-09-17","cited_by_count":1,"area":"vla"},{"id":"2609.20791","title":"StageGuard: Learning Stage Transitions for Long-Horizon Robot Tasks via Agentic Distillation","created":"2026-09-17","cited_by_count":1,"area":"agents"},{"id":"2609.21755","title":"ECG Mirage: Revealing and Mitigating the Underutilisation of ECGs in Vision-Language Models for Clinical Prediction","created":"2026-09-18","cited_by_count":1,"area":null},{"id":"2609.23417","title":"Omni2Web: Benchmarking Audiovisual Website Development","created":"2026-09-20","cited_by_count":1,"area":null},{"id":"2609.25652","title":"GameDirector: Decoupling Gameplay Logic from Rendering for Player-Configurable Game World Models","created":"2026-09-22","cited_by_count":1,"area":"wm"},{"id":"2609.29287","title":"AEGIS: Audio Endogenous Guarding via Internal Signals Against Large Audio-Language Model Jailbreaks","created":"2026-09-24","cited_by_count":1,"area":"safety"},{"id":"2609.32248","title":"Anytime-Valid LLM Leaderboards via Benchmark-weighted and Block-Factorized e-Processes","created":"2026-09-26","cited_by_count":1,"area":null},{"id":"2609.32610","title":"KV-Lingo: Learning KV-Cache Translators with Distillation","created":"2026-09-26","cited_by_count":1,"area":"inference"},{"id":"2609.36777","title":"Code4Scene: Benchmarking Coding Agents for Constructing and Editing 3D Scenes","created":"2026-09-29","cited_by_count":1,"area":"agents"},{"id":"2609.37988","title":"KV-Kaizen: Learning Context-Adaptive Cache Compression Choices","created":"2026-09-29","cited_by_count":1,"area":"inference"},{"id":"2610.04184","title":"EvalResearchBench: Can AI Agents Design Their Own Evaluations?","created":"2026-10-03","cited_by_count":1,"area":null},{"id":"2607.25431","title":"CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents","created":"2026-07-28","cited_by_count":0,"area":"agents"},{"id":"2607.26326","title":"Seeing or Knowing? Visual Context Sensitivity in Multimodal Large Language Models","created":"2026-07-28","cited_by_count":0,"area":"mllm"},{"id":"2608.01112","title":"Fighting Fire with Fire: On the Feasibility of Protecting Exercises Against AI Cheating","created":"2026-08-02","cited_by_count":0,"area":null},{"id":"2608.01409","title":"When Retrieval Helps and Distracts: Evaluating Evidence-Generating LLMs for Biomedical Claim Verification","created":"2026-08-02","cited_by_count":0,"area":null},{"id":"2608.06115","title":"Mind the Gaps: Mixture-of-Minds for Human Simulation","created":"2026-08-06","cited_by_count":0,"area":null},{"id":"2608.06758","title":"Stockmark-Nemotron-3-Nano-Omni-JapanDocReader: Structured Document Parsing via Capability Injection and Forgetting Control","created":"2026-08-07","cited_by_count":0,"area":null},{"id":"2608.08939","title":"Not an A11y: How Android Accessibility Exposes Mobile AI Agents to Indirect Prompt Injection","created":"2026-08-09","cited_by_count":0,"area":"safety"},{"id":"2608.09593","title":"MADBench: A Benchmark for Modality-Aware Audio Deepfake Detection","created":"2026-08-10","cited_by_count":0,"area":null},{"id":"2608.11705","title":"Making Your LLMs More Objective: Stabilizing LLM Safety Behavior Across Traits with Trait-Invariant Safety Tuning","created":"2026-08-12","cited_by_count":0,"area":"alignment"},{"id":"2608.12585","title":"Reasoning Jury: Multi-Model Consensus for Evaluating Reasoning Traces","created":"2026-08-12","cited_by_count":0,"area":"lr"},{"id":"2608.14927","title":"LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks","created":"2026-08-14","cited_by_count":0,"area":"lr"},{"id":"2608.14999","title":"RamseyGadgets: A Graph Construction Dataset for LLMs","created":"2026-08-15","cited_by_count":0,"area":null},{"id":"2608.16970","title":"Probing the Prefill: Detecting Code Vulnerabilities via Latent Activations","created":"2026-08-17","cited_by_count":0,"area":null},{"id":"2608.17687","title":"Mixture-of-Expert Blocks Contain Strong Hallucination Detection Signals","created":"2026-08-18","cited_by_count":0,"area":"moe"},{"id":"2608.17895","title":"BEAR-Bench: A Bilingual Enterprise and Academic Reasoning Benchmark for Multimodal Models","created":"2026-08-18","cited_by_count":0,"area":"lr"},{"id":"2608.18011","title":"The IOL-AI Challenge: An Open Challenge towards Advancing Linguistic Reasoning","created":"2026-08-18","cited_by_count":0,"area":"lr"},{"id":"2608.19863","title":"Listening Forward: Next Patch Embedding Prediction Enables Scalable Audio Learners","created":"2026-08-20","cited_by_count":0,"area":null},{"id":"2608.28384","title":"MAP: A Benchmark on Multimodal Accessibility Planning for Real World Places","created":"2026-08-28","cited_by_count":0,"area":null},{"id":"2608.29120","title":"HEAR Who Said What: Unlocking Speaker-Attributed Reasoning via Counterfactual Voice Grounding","created":"2026-08-29","cited_by_count":0,"area":"lr"},{"id":"2608.30068","title":"TAKE 85: Testing Audiovisual filmmaKer's intEnt across 85 Hours of Film","created":"2026-08-30","cited_by_count":0,"area":null},{"id":"2608.30703","title":"SingProbe Technical Report","created":"2026-08-31","cited_by_count":0,"area":"safety"},{"id":"2608.30719","title":"Mind the Gap: Theory-of-Mind-Grounded Friction for Epistemic Alignment","created":"2026-08-31","cited_by_count":0,"area":"alignment"},{"id":"2609.00590","title":"CRAFT: Fine-Tuning Pre-hoc Explainability in AI-native 6G RAN","created":"2026-09-01","cited_by_count":0,"area":"lr"},{"id":"2609.01657","title":"NeoMME: A Single-Tower Multimodal-Native Multilingual Foundation Encoder for Efficient Fine-Tuning and Inference","created":"2026-08-31","cited_by_count":0,"area":"mllm"},{"id":"2609.02573","title":"Deeply Interleaved Text-Image Contexts for Multimodal LLMs Assessment","created":"2026-09-02","cited_by_count":0,"area":"mllm"},{"id":"2609.03526","title":"CulturalMenuBench: Probing the Knowledge-Application Gap in Multimodal Culinary Reasoning","created":"2026-09-03","cited_by_count":0,"area":"lr"},{"id":"2609.03611","title":"FailBench: How Reliable are VLMs at Judging Robot Task Success?","created":"2026-09-03","cited_by_count":0,"area":"mllm"},{"id":"2609.04753","title":"Beneath the Surface of Chains-of-Thought: A Mechanistic Interpretation of Reasoning Operations in LLMs","created":"2026-09-04","cited_by_count":0,"area":"lr"},{"id":"2609.05517","title":"Emergent Goal-Directed Attention in Large Vision-Language Models","created":"2026-08-31","cited_by_count":0,"area":"mllm"},{"id":"2609.05535","title":"Beyond the Verdict: Evidence-Aligned Evaluation of Visual Prompt-Injection Guardrails","created":"2026-09-02","cited_by_count":0,"area":"mllm"},{"id":"2609.05797","title":"Safety Monitors Mostly Catch What the Model Already Refuses","created":"2026-09-05","cited_by_count":0,"area":"safety"},{"id":"2609.06880","title":"Contextual Observer Grounding: Evaluating Situated Spatial Reasoning in Vision-Language Models","created":"2026-09-07","cited_by_count":0,"area":"lr"},{"id":"2609.06970","title":"CantoneseLLM v2: Reasoning in a Low-Resource Language","created":"2026-09-07","cited_by_count":0,"area":"lr"},{"id":"2609.07131","title":"AgentLeak: Cloning Stronger LLM Agent Capabilities onto Weaker Agents Beyond Skill Stealing","created":"2026-09-07","cited_by_count":0,"area":"agents"},{"id":"2609.09054","title":"Training-Free Task Vectors for LLM Behavioral Control","created":"2026-09-08","cited_by_count":0,"area":null},{"id":"2609.09417","title":"Vision-language models know more about agriculture than they show and rubric-grounded verifications close the gap","created":"2026-09-08","cited_by_count":0,"area":"mllm"},{"id":"2609.09528","title":"MotionBlind: Probing the Illusion of Motion Understanding in Video-LLMs","created":"2026-09-08","cited_by_count":0,"area":"wm"},{"id":"2609.11061","title":"Fork Where the Model Changes Its Mind: Belief-Shift Branching for Tree-Structured Reinforcement Learning","created":"2026-09-10","cited_by_count":0,"area":"lr"},{"id":"2609.11209","title":"REVA: Reusable Evidence View Aggregation for Context-Efficient RAG Serving","created":"2026-09-10","cited_by_count":0,"area":"rag"},{"id":"2609.12190","title":"Retrieval-Augmented Generation for Scientific Code Understanding","created":"2026-09-10","cited_by_count":0,"area":"rag"},{"id":"2609.12509","title":"ChitraMiti: Benchmarking Visual Grounding and Modality Reliance in Bengali Geometric Reasoning","created":"2026-09-11","cited_by_count":0,"area":"lr"},{"id":"2609.12663","title":"What Did the MLLM Hear? Token-Level Spectro-Temporal Grounding for Audio MLLM Explainability","created":"2026-09-11","cited_by_count":0,"area":"mllm"}],"next_cursor":"100"}