Klear-Reasoner: Advancing Reasoning Capability via Gradient-Preserving Clipping Policy Optimization
Paper
• 2508.07629 • Published
• 43
Reasoning Papers
-
-
Less Is More: Training-Free Sparse Attention with Global Locality for Efficient Reasoning
Paper • 2508.07101 • Published • 14 -
Compressing Chain-of-Thought in LLMs via Step Entropy
Paper • 2508.03346 • Published • 8 -
Train Long, Think Short: Curriculum Learning for Efficient Reasoning
Paper • 2508.08940 • Published • 27 -
Sample More to Think Less: Group Filtered Policy Optimization for Concise Reasoning
Paper • 2508.09726 • Published • 15 -
Pass@k Training for Adaptively Balancing Exploration and Exploitation of Large Reasoning Models
Paper • 2508.10751 • Published • 29 -
Beyond Solving Math Quiz: Evaluating the Ability of Large Reasoning Models to Ask for Information
Paper • 2508.11252 • Published • 3 -
Deep Think with Confidence
Paper • 2508.15260 • Published • 90 -
Beyond Pass@1: Self-Play with Variational Problem Synthesis Sustains RLVR
Paper • 2508.14029 • Published • 118 -
CARFT: Boosting LLM Reasoning via Contrastive Learning with Annotated Chain-of-Thought-based Reinforced Fine-Tuning
Paper • 2508.15868 • Published • 3 -
Breaking the Exploration Bottleneck: Rubric-Scaffolded Reinforcement Learning for General LLM Reasoning
Paper • 2508.16949 • Published • 24 -
TreePO: Bridging the Gap of Policy Optimization and Efficacy and Inference Efficiency with Heuristic Tree-based Modeling
Paper • 2508.17445 • Published • 80 -
ThinkDial: An Open Recipe for Controlling Reasoning Effort in Large Language Models
Paper • 2508.18773 • Published • 16 -
StepWiser: Stepwise Generative Judges for Wiser Reasoning
Paper • 2508.19229 • Published • 20 -
Reasoning Vectors: Transferring Chain-of-Thought Capabilities via Task Arithmetic
Paper • 2509.01363 • Published • 59 -
Implicit Actor Critic Coupling via a Supervised Learning Framework for RLVR
Paper • 2509.02522 • Published • 26 -
Loong: Synthesize Long Chain-of-Thoughts at Scale through Verifiers
Paper • 2509.03059 • Published • 25 -
Reverse-Engineered Reasoning for Open-Ended Generation
Paper • 2509.06160 • Published • 149 -
Parallel-R1: Towards Parallel Thinking via Reinforcement Learning
Paper • 2509.07980 • Published • 105 -
Staying in the Sweet Spot: Responsive Reasoning Evolution via Capability-Adaptive Hint Scaffolding
Paper • 2509.06923 • Published • 22 -
Emergent Hierarchical Reasoning in LLMs through Reinforcement Learning
Paper • 2509.03646 • Published • 33 -
A Survey of Reinforcement Learning for Large Reasoning Models
Paper • 2509.08827 • Published • 190 -
The Majority is not always right: RL training for solution aggregation
Paper • 2509.06870 • Published • 15 -
The Choice of Divergence: A Neglected Key to Mitigating Diversity Collapse in Reinforcement Learning with Verifiable Reward
Paper • 2509.07430 • Published • 3 -
Reasoning-Aware GRPO using Process Mining
Paper • 2510.25065 • Published • 42 -
Scaling Latent Reasoning via Looped Language Models
Paper • 2510.25741 • Published • 229 -
FAPO: Flawed-Aware Policy Optimization for Efficient and Reliable Reasoning
Paper • 2510.22543 • Published • 14 -
Supervised Reinforcement Learning: From Expert Trajectories to Step-wise Reasoning
Paper • 2510.25992 • Published • 48 -
SemCoT: Accelerating Chain-of-Thought Reasoning through Semantically-Aligned Implicit Tokens
Paper • 2510.24940 • Published • 18 -
MR-Align: Meta-Reasoning Informed Factuality Alignment for Large Reasoning Models
Paper • 2510.24794 • Published • 32 -
Data-Efficient RLVR via Off-Policy Influence Guidance
Paper • 2510.26491 • Published • 11 -
Black-Box On-Policy Distillation of Large Language Models
Paper • 2511.10643 • Published • 52 -
Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models
Paper • 2511.08577 • Published • 108 -
DeepSeekMath-V2: Towards Self-Verifiable Mathematical Reasoning
Paper • 2511.22570 • Published • 91 -
REFLEX: Self-Refining Explainable Fact-Checking via Disentangling Truth into Style and Substance
Paper • 2511.20233 • Published • 3 -
Arbitrage: Efficient Reasoning via Advantage-Aware Speculation
Paper • 2512.05033 • Published • 17 -
LYNX: Learning Dynamic Exits for Confidence-Controlled Reasoning
Paper • 2512.05325 • Published • 4 -
Nemotron-Math: Efficient Long-Context Distillation of Mathematical Reasoning from Multi-Mode Supervision
Paper • 2512.15489 • Published • 10 -
Fantastic Reasoning Behaviors and Where to Find Them: Unsupervised Discovery of the Reasoning Process
Paper • 2512.23988 • Published • 18 -
RelayLLM: Efficient Reasoning via Collaborative Decoding
Paper • 2601.05167 • Published • 31 -
DiffCoT: Diffusion-styled Chain-of-Thought Reasoning in LLMs
Paper • 2601.03559 • Published • 14 -
The Molecular Structure of Thought: Mapping the Topology of Long Chain-of-Thought Reasoning
Paper • 2601.06002 • Published • 56 -
Where Did This Sentence Come From? Tracing Provenance in LLM Reasoning Distillation
Paper • 2512.20908 • Published • 29 -
Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning
Paper • 2601.09088 • Published • 63 -
Which Reasoning Trajectories Teach Students to Reason Better? A Simple Metric of Informative Alignment
Paper • 2601.14249 • Published • 12 -
Teaching Models to Teach Themselves: Reasoning at the Edge of Learnability
Paper • 2601.18778 • Published • 40 -
Harder Is Better: Boosting Mathematical Reasoning via Difficulty-Aware GRPO and Multi-Aspect Question Reformulation
Paper • 2601.20614 • Published • 119 -
DenseGRPO: From Sparse to Dense Reward for Flow Matching Model Alignment
Paper • 2601.20218 • Published • 15 -
Memorization Dynamics in Knowledge Distillation for Language Models
Paper • 2601.15394 • Published • 3