{"total":137,"rows":[{"id":"2606.15007","title":"Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning","created":"2026-06-12","cited_by_count":42,"area":"lr"},{"id":"2605.21606","title":"When Are Teacher Tokens Reliable? Position-Weighted On-Policy Self-Distillation for Reasoning","created":"2026-05-20","cited_by_count":30,"area":"distill"},{"id":"2607.13399","title":"Demystifying On-Policy Distillation: Roles, Pathologies, and Regulations","created":"2026-07-15","cited_by_count":19,"area":"distill"},{"id":"2608.04794","title":"Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation","created":"2026-08-05","cited_by_count":18,"area":"distill"},{"id":"2605.12400","title":"OGLS-SD: On-Policy Self-Distillation with Outcome-Guided Logit Steering for LLM Reasoning","created":"2026-05-12","cited_by_count":16,"area":"lr"},{"id":"2607.15161","title":"On-Policy Delta Distillation","created":"2026-07-16","cited_by_count":15,"area":"alignment"},{"id":"2605.11609","title":"Anti-Self-Distillation for Reasoning RL via Pointwise Mutual Information","created":"2026-05-12","cited_by_count":14,"area":"lr"},{"id":"2605.30833","title":"Your Teacher Can't Help You Here: Combating Supervision Fidelity Decay in On-Policy Distillation","created":"2026-05-29","cited_by_count":13,"area":"lr"},{"id":"2605.09253","title":"Cornerstones or Stumbling Blocks? Deciphering the Rock Tokens in On-Policy Distillation","created":"2026-05-10","cited_by_count":11,"area":"distill"},{"id":"2606.06712","title":"Data-Efficient Autoregressive-to-Diffusion Language Models via On-Policy Distillation","created":"2026-06-04","cited_by_count":11,"area":"diffusion"},{"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":"2606.00206","title":"Quantized Reasoning Models Think They Need to Think Longer, but They Do Not","created":"2026-05-29","cited_by_count":10,"area":"lr"},{"id":"2606.12370","title":"Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling","created":"2026-06-10","cited_by_count":9,"area":null},{"id":"2606.01476","title":"OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification","created":"2026-05-31","cited_by_count":7,"area":"lr"},{"id":"2606.23104","title":"ReNIO: Reweighting Negative Trajectory Importance for LLM On-Policy Distillation","created":"2026-06-22","cited_by_count":7,"area":"lr"},{"id":"2608.04001","title":"Test-Time Scaling in Reasoning LLMs: Inference Regimes, Evaluation, and Reproducibility","created":"2026-08-04","cited_by_count":6,"area":"lr"},{"id":"2608.27513","title":"DAMP: Decay-Aware Mixed-Precision Recurrent-State Quantization","created":"2026-08-27","cited_by_count":6,"area":"inference"},{"id":"2607.10805","title":"Diagnosing and Mitigating Thinking Collapse in On-Policy Self-Distillation","created":"2026-07-12","cited_by_count":5,"area":"lr"},{"id":"2608.12486","title":"DIVE: Unlocking Self-Improvement in Frozen Language Models Through Diversity-Driven Skill Evolution","created":"2026-08-12","cited_by_count":5,"area":"synth"},{"id":"2608.31075","title":"Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence","created":"2026-08-31","cited_by_count":5,"area":"lr"},{"id":"2605.05365","title":"ZAYA1-8B Technical Report","created":"2026-05-06","cited_by_count":4,"area":"lr"},{"id":"2605.20315","title":"Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs","created":"2026-05-19","cited_by_count":4,"area":"agents"},{"id":"2608.13505","title":"Intern-S2-Preview: Scientific Agentic Foundation Model","created":"2026-08-13","cited_by_count":4,"area":"agents"},{"id":"2605.21468","title":"You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectories","created":"2026-05-20","cited_by_count":3,"area":"lr"},{"id":"2606.02398","title":"A Local Perturbation Theory for Cross-Domain Interference and Recovery in Multi-Domain RL","created":"2026-06-01","cited_by_count":3,"area":"lr"},{"id":"2606.04883","title":"Optimizing the Cost-Quality Tradeoff of Agentic Theorem Provers in Lean","created":"2026-06-03","cited_by_count":3,"area":"agents"},{"id":"2606.18284","title":"Breaking the Solver Bottleneck: Training Task Generators at the Learnable Frontier","created":"2026-06-10","cited_by_count":3,"area":"lr"},{"id":"2608.01667","title":"TCPO: Turn-Level Credit Policy Optimization","created":"2026-08-03","cited_by_count":3,"area":"lr"},{"id":"2605.05873","title":"CITE: Anytime-Valid Statistical Inference in LLM Self-Consistency","created":"2026-05-07","cited_by_count":2,"area":"lr"},{"id":"2605.06326","title":"Teaching Thinking Models to Reason with Tools: A Full-Pipeline Recipe for Tool-Integrated Reasoning","created":"2026-05-07","cited_by_count":2,"area":"lr"},{"id":"2605.27030","title":"Share More, Search Less: Collaborative Parallel Thinking for Efficient Test-Time Scaling","created":"2026-05-26","cited_by_count":2,"area":"lr"},{"id":"2605.28003","title":"ResearchMath-14K: Scaling Research-Level Mathematics via Agents","created":"2026-05-27","cited_by_count":2,"area":"agents"},{"id":"2605.29310","title":"Rubric-Guided Process Reward for Stepwise Model Routing","created":"2026-05-28","cited_by_count":2,"area":"lr"},{"id":"2606.03102","title":"Small RL Controller, Large Language Model: RL-Guided Adaptive Sampling for Test-Time Scaling","created":"2026-06-02","cited_by_count":2,"area":"lr"},{"id":"2607.28582","title":"$\\beta$-OPSD: Deriving with Policy Optimization, Training with Self-Distillation","created":"2026-07-30","cited_by_count":2,"area":"lr"},{"id":"2608.01662","title":"LongCat Sparse Attention: Taming the Lightning via Streaming-aware Hierarchical Cross-Layer Indexing","created":"2026-08-03","cited_by_count":2,"area":null},{"id":"2608.24987","title":"D$^3$-MOPD: Dynamic Domain ScheDuling for Efficient Multi-Teacher Distillation","created":"2026-08-25","cited_by_count":2,"area":"distill"},{"id":"2608.27448","title":"TTPO: Test-Time Policy Optimization","created":"2026-08-27","cited_by_count":2,"area":"lr"},{"id":"2608.30386","title":"DASC: Decay-Aware State Compression for Hybrid Linear-Attention Serving","created":"2026-08-31","cited_by_count":2,"area":null},{"id":"2609.04010","title":"Unlocking Lossless Speedups in LLMs via Discrete Diffusion","created":"2026-09-03","cited_by_count":2,"area":"diffusion"},{"id":"2609.08798","title":"Eliciting Weak-to-Strong Generalization with On-Policy Reverse Distillation","created":"2026-09-08","cited_by_count":2,"area":"distill"},{"id":"2609.11699","title":"Negative Self-Distillation: Learning to Reason by Avoiding Flaws","created":"2026-09-10","cited_by_count":2,"area":"lr"},{"id":"2609.24432","title":"1% of Tokens Can Be Enough: On Gradient Estimation in On-Policy Distillation","created":"2026-09-21","cited_by_count":2,"area":"distill"},{"id":"2605.27712","title":"Prefix-Safe Bayesian Belief Tracking for LLM Reasoning Reliability:Separating Calibration from Ranking","created":"2026-05-26","cited_by_count":1,"area":"lr"},{"id":"2605.28150","title":"Off-Policy Learning to Reason Works Because It Is More Pessimistic Than You Think","created":"2026-05-27","cited_by_count":1,"area":"lr"},{"id":"2606.19605","title":"FAPO: Fully Automated Prompt Optimization of Multi-Step LLM Pipelines","created":"2026-06-17","cited_by_count":1,"area":null},{"id":"2607.02942","title":"Serving Agentic Workflows with a Physical-Plan Compiler and Adaptive Runtime","created":"2026-07-03","cited_by_count":1,"area":"agents"},{"id":"2607.10114","title":"Cost of Reasoning in non-English Languages: A Case Study on Japanese","created":"2026-07-11","cited_by_count":1,"area":"lr"},{"id":"2607.10836","title":"Route, Communicate, and Reason: Gated Routing and Adaptive Depth for Efficient Multi-Agent Reasoning","created":"2026-07-12","cited_by_count":1,"area":null},{"id":"2607.27600","title":"Back from the Future: Key-Value Cache Management by Counter-Causal Surprise","created":"2026-07-30","cited_by_count":1,"area":"inference"},{"id":"2608.02149","title":"Beyond the Mean: Multi-Moment Policy Optimization for LLM Reasoning","created":"2026-08-03","cited_by_count":1,"area":"lr"},{"id":"2608.11994","title":"Claim-Level Reliability Assessment for Efficient Test-Time Reasoning","created":"2026-08-12","cited_by_count":1,"area":"lr"},{"id":"2608.16425","title":"ParaTempo: Efficient Parallel Reasoning via Temporal Confidence","created":"2026-08-17","cited_by_count":1,"area":"lr"},{"id":"2608.17941","title":"Efficient RLVR Scheduling via Graph-Structured Online Difficulty Estimation","created":"2026-08-18","cited_by_count":1,"area":"lr"},{"id":"2608.20634","title":"AgentMercury: Your Agent Can Synthesize Verifiable Environments for Business Scenarios at scale","created":"2026-08-21","cited_by_count":1,"area":null},{"id":"2609.13443","title":"Learning to Solve Hard Problems in RL for LLMs by Never Giving Up","created":"2026-09-11","cited_by_count":1,"area":null},{"id":"2609.17890","title":"OBC-Prune: Outcome-Based Calibration for Large Reasoning Model Pruning","created":"2026-09-15","cited_by_count":1,"area":"lr"},{"id":"2609.21561","title":"On Repulsive and Attractive Teachers: Separating Correctness from Behavior in Self-Distillation","created":"2026-09-18","cited_by_count":1,"area":"distill"},{"id":"2609.26637","title":"Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models","created":"2026-09-22","cited_by_count":1,"area":"lr"},{"id":"2610.03039","title":"HyperThink: Text-to-Parameter Hypernetworks for Efficient Reasoning","created":"2026-10-02","cited_by_count":1,"area":"lr"},{"id":"2605.11905","title":"Rethinking Supervision Granularity: Segment-Level Learning for LLM-Based Theorem Proving","created":"2026-05-12","cited_by_count":0,"area":"lr"},{"id":"2606.05563","title":"SoCRATES: Towards Reliable Automated Evaluation of Proactive LLM Mediation across Domains and Socio-cognitive Variations","created":"2026-06-04","cited_by_count":0,"area":null},{"id":"2606.07515","title":"How reliable are LLMs when it comes to playing dice?","created":"2026-06-05","cited_by_count":0,"area":"lr"},{"id":"2606.16154","title":"A Gradient Perspective on RLVR Stability and Winner Advantage Policy Optimization","created":"2026-06-15","cited_by_count":0,"area":"lr"},{"id":"2606.24084","title":"Blockwise Policy-Drift Gating for On-Policy Distillation","created":"2026-06-23","cited_by_count":0,"area":"distill"},{"id":"2606.25451","title":"Learning with a Single Rollout via Pass@k Policy Optimization","created":"2026-06-24","cited_by_count":0,"area":null},{"id":"2607.08017","title":"Can We Trust LLM's Logic? Quantifying Uncertainty, Coherence, and Robustness via a Graph-Based Framework","created":"2026-07-09","cited_by_count":0,"area":"lr"},{"id":"2607.08317","title":"Blind-Spots-Bench: Evaluating Blind Spots in Multimodal Models","created":"2026-07-09","cited_by_count":0,"area":null},{"id":"2607.08837","title":"Prompt-Driven Exploration: Language as an Exploration Space for VLA Reinforcement Learning","created":"2026-07-09","cited_by_count":0,"area":null},{"id":"2607.11444","title":"UMoE:Unlocking Every Expert in Domain-Specific Training","created":"2026-07-13","cited_by_count":0,"area":"moe"},{"id":"2607.11849","title":"AdvancedMathBench: A Benchmark Suite for Advanced Mathematical Proof Generation and Verification","created":"2026-07-13","cited_by_count":0,"area":null},{"id":"2607.27888","title":"Not All Tokens Deserve Equal Credit: Counterfactual Sensitivity Credit Reallocation for Long-CoT Reasoning","created":"2026-07-30","cited_by_count":0,"area":"lr"},{"id":"2608.00013","title":"What Transfers from Text to Vision? Capability Scaling Laws and Transfer Dynamics for VLMs","created":"2026-06-24","cited_by_count":0,"area":"mllm"},{"id":"2608.00533","title":"Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian Languages","created":"2026-08-01","cited_by_count":0,"area":"lr"},{"id":"2608.04505","title":"K-EXAONE 2.0 Technical Report","created":"2026-08-05","cited_by_count":0,"area":"moe"},{"id":"2608.06933","title":"Ask-E: An Environment for Calibrated Question Generation","created":"2026-08-07","cited_by_count":0,"area":null},{"id":"2608.09119","title":"Motif 3: Technical Report","created":"2026-08-10","cited_by_count":0,"area":"moe"},{"id":"2608.12585","title":"Reasoning Jury: Multi-Model Consensus for Evaluating Reasoning Traces","created":"2026-08-12","cited_by_count":0,"area":"lr"},{"id":"2608.21871","title":"The Chase Is the Curriculum, the Capture Anchors the Credit: Pursuit-Evasion Self-Play for Zero-Data LLM Reasoning","created":"2026-08-22","cited_by_count":0,"area":"lr"},{"id":"2608.23256","title":"Is Next-Chunk Reasoning RL Really Better than SFT? Revisiting Training Strategies under no-CoT Data","created":"2026-08-24","cited_by_count":0,"area":"lr"},{"id":"2608.26684","title":"Reason in the Words You Speak: Idiolectal Paraphrasing Off-Policy Traces for Reasoning Distillation in VideoLLMs","created":"2026-08-27","cited_by_count":0,"area":"lr"},{"id":"2608.30181","title":"A.X K2 Technical Report","created":"2026-08-31","cited_by_count":0,"area":"agents"},{"id":"2609.00062","title":"RePro: Proof-Verified Benchmark Rewriting for Reliable Evaluation of LLM Mathematical Problem Solving","created":"2026-08-30","cited_by_count":0,"area":null},{"id":"2609.08966","title":"Good Pretraining, Bad SFT: Checkpoint Quality Across the Training Stack","created":"2026-09-08","cited_by_count":0,"area":null},{"id":"2609.10154","title":"CompassOPD: Cross-Family On-Policy Distillation via Within-Family Likelihood Shifts","created":"2026-09-09","cited_by_count":0,"area":"distill"},{"id":"2609.10712","title":"An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics","created":"2026-09-09","cited_by_count":0,"area":null},{"id":"2609.12419","title":"MInTRL: Off-policy Intervention can boost On-policy RL","created":"2026-09-11","cited_by_count":0,"area":null},{"id":"2609.12578","title":"From Collaboration to Capability: Internalizing Routed LLM Experts into Compact Reasoners","created":"2026-09-11","cited_by_count":0,"area":"lr"},{"id":"2609.23989","title":"ACLArena: Agent Continue Learning in Multi-stage Post-training","created":"2026-09-21","cited_by_count":0,"area":null},{"id":"2609.27374","title":"Planned Test-Time Scaling with Coordinated Reasoning Paths","created":"2026-09-23","cited_by_count":0,"area":"lr"},{"id":"2609.28560","title":"Speculative Evaluation of Stochastic LLMs","created":"2026-09-23","cited_by_count":0,"area":null},{"id":"2609.31009","title":"G$^2$PTQ: Improving LLM Post-Training Quantization with Generalized Gradient Compensation","created":"2026-09-25","cited_by_count":0,"area":null},{"id":"2609.31619","title":"Learning to Stop without Learning to Stop: Self-Supervised Confidence Training Improves Reasoning Efficiency","created":"2026-09-25","cited_by_count":0,"area":"lr"},{"id":"2609.32470","title":"On the Pitfalls of Verbalized Confidence Priors for Calibrating Large Reasoning Models","created":"2026-09-26","cited_by_count":0,"area":"lr"},{"id":"2609.33271","title":"Next Thoughts Are Distributions: Generative Autoregressive Reasoning in the Latent Space","created":"2026-09-27","cited_by_count":0,"area":null},{"id":"2609.33559","title":"Fine Until Fine-Tuned: Repeated Solutions Make Reasoning Fragile","created":"2026-09-27","cited_by_count":0,"area":null},{"id":"2609.33642","title":"Learning to Learn from Context: Synthetic Training from Perturbed Public Documents","created":"2026-09-27","cited_by_count":0,"area":null},{"id":"2609.34225","title":"USA: Update-aware SAM for Cross-domain On-Policy Disitllation of Language Agents","created":"2026-09-28","cited_by_count":0,"area":"agents"},{"id":"2609.34327","title":"Knowing When Thinking Is Not Enough: Teaching Small Reasoning Models to Reason Beyond Their Parametric Knowledge","created":"2026-09-28","cited_by_count":0,"area":"lr"},{"id":"2609.34848","title":"Can We Trust the Teacher? Decoupled Credit Direction-Magnitude for Self-Distillation","created":"2026-09-28","cited_by_count":0,"area":"distill"}],"next_cursor":"100"}