Reasoning and the "physics" of language models — page 3

• Reasoning on the Simplex: Geometric Fixed-Point Models
  arxiv:2609.33540 · 2026-09-27 · text available
• Reinforcement Learning for Reasoning in Large Language Models with One Training Example
  arxiv:2504.20571 · 2025-04-29 · text available
• Reinforcing General Reasoning without Verifiers
  arxiv:2505.21493 · 2025-05-27 · text available
• Reliable Chain-of-Thought via Prefix Consistency
  arxiv:2605.07654 · 2026-05-08 · text available
• rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking
  arxiv:2501.04519 · 2025-01-08 · text available
• s1: Simple test-time scaling
  arxiv:2501.19393 · 2025-01-31 · text available
• Scaling of Search and Learning: A Roadmap to Reproduce o1 from Reinforcement Learning Perspective
  arxiv:2412.14135 · 2024-12-18 · text available
• Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
  arxiv:2502.05171 · 2025-02-07 · text available
• Self-Consistency Improves Chain of Thought Reasoning in Language Models
  arxiv:2203.11171 · 2022-03-21 · text available
• SIM-CoT: Supervised Implicit Chain-of-Thought
  arxiv:2509.20317 · 2025-09-24 · text available
• Solve the Loop: Attractor Models for Language and Reasoning
  arxiv:2605.12466 · 2026-05-12 · text available
• Spurious Rewards: Rethinking Training Signals in RLVR
  arxiv:2506.10947 · 2025-06-12 · text available
• The Illusion of Latent Generalization: Bi-directionality and the Reversal Curse
  arxiv:2604.04943 · 2026-03-13 · text available
• The Lookahead Limitation: Why Multi-Operand Addition is Hard for LLMs
  arxiv:2502.19981 · 2025-02-27 · text available
• The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"
  arxiv:2309.12288 · 2023-09-21 · text available
• The Truth is in There: Improving Reasoning in Language Models with Layer-Selective Rank Reduction
  arxiv:2312.13558 · 2023-12-21 · text available
• The Unreasonable Effectiveness of Entropy Minimization in LLM Reasoning
  arxiv:2505.15134 · 2025-05-21 · text available
• Training Large Language Models to Reason in a Continuous Latent Space
  arxiv:2412.06769 · 2024-12-09 · text available
• Transcendence: Generative Models Can Outperform The Experts That Train Them
  arxiv:2406.11741 · 2024-06-17 · text available
• Universal Reasoning Model
  arxiv:2512.14693 · 2025-12-16 · text available
• Universal Transformers Need Memory: Depth-State Trade-offs in Adaptive Recursive Reasoning
  arxiv:2604.21999 · 2026-04-23 · text available
• Why Can't Transformers Learn Multiplication? Reverse-Engineering Reveals Long-Range Dependency Pitfalls
  arxiv:2510.00184 · 2025-09-30 · text available
• Why mathematics is set to be revolutionized by AI
  doi:10.1038/d41586-024-01413-w · 2024-05-14 · text available