{"total":67,"papers":[{"id":"arxiv:2608.09888","title":"BDH-CQ: In-Context Learning with Recurrent Latent Reasoning","date":"2026-08-10","section":"Post-training","quality":-0.339191,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/f0c77a8a0f50254e9efdf9a511e08fc2a40677bfa029864bdd0a9cbce561c666/","source_url":"https://arxiv.org/abs/2608.09888","text_available":true},{"id":"arxiv:2506.01939","title":"Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning","date":"2025-06-02","section":"Post-training","quality":0.275648,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/e0f4a55bca612bce5895b6b98f89b0efc14cbd8cf45946a7743347cf6daee1a1/","source_url":"https://arxiv.org/abs/2506.01939","text_available":true},{"id":"arxiv:2509.14234","title":"Compute as Teacher: Turning Inference Compute Into Reference-Free Supervision","date":"2025-09-17","section":"Post-training","quality":0.066281,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/a6f1443a7e55dd6421fdbe43e65e19c1962a19b135d737fd67f414d3324ce85e/","source_url":"https://arxiv.org/abs/2509.14234","text_available":true},{"id":"arxiv:2506.03106","title":"Critique-GRPO: Advancing LLM Reasoning with Natural Language and Numerical Feedback","date":"2025-06-03","section":"Post-training","quality":0.423423,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/b16b5772846ee30548a2d2c9cce1833ca3af6beb01c4a58f2dd5a69d809d55d4/","source_url":"https://arxiv.org/abs/2506.03106","text_available":true},{"id":"arxiv:2506.06632","title":"Curriculum Reinforcement Learning from Easy to Hard Tasks Improves LLM Reasoning","date":"2025-06-07","section":"Post-training","quality":0.051441,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/79fda9609f9b88bb71f71f531b63ea6b5237d1c889b7ae59e9ca07c12dbe70d1/","source_url":"https://arxiv.org/abs/2506.06632","text_available":true},{"id":"arxiv:2503.14476","title":"DAPO: An Open-Source LLM Reinforcement Learning System at Scale","date":"2025-03-18","section":"Post-training","quality":-0.315765,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/548e934b912ef84691ace63bf1cfd670e7b0257d9d6986dde926f569e7b2f8b7/","source_url":"https://arxiv.org/abs/2503.14476","text_available":true},{"id":"arxiv:2402.03300","title":"DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models","date":"2024-02-05","section":"Post-training","quality":0.322538,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/636fb871904bc06f742cfdea069796f9e9665f95bd58dc63260eaff5154ef985/","source_url":"https://arxiv.org/abs/2402.03300","text_available":true},{"id":"arxiv:2504.13837","title":"Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?","date":"2025-04-18","section":"Post-training","quality":0.022969,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/b90f70269aeb02311d74ffaadf47e139683424a47118dc18619000ba72174d17/","source_url":"https://arxiv.org/abs/2504.13837","text_available":true},{"id":"arxiv:2509.03646","title":"Emergent Hierarchical Reasoning in LLMs through Reinforcement Learning","date":"2025-09-03","section":"Post-training","quality":-0.042926,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/683657af0d699b7d854ce84d95bcd53c12e152aa738c602d0982cc0d11009f08/","source_url":"https://arxiv.org/abs/2509.03646","text_available":true},{"id":"arxiv:2512.00499","title":"ESPO: Entropy Importance Sampling Policy Optimization","date":"2025-11-29","section":"Post-training","quality":0.192562,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/342b84df2bf40f64c57ea7839579c1f337374ad9c28b16ce7d477c42e24c67b4/","source_url":"https://arxiv.org/abs/2512.00499","text_available":true},{"id":"arxiv:2601.20861","title":"Evolutionary Strategies lead to Catastrophic Forgetting in LLMs","date":"2026-01-28","section":"Post-training","quality":-0.530464,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/4178758b0cb3ebc95f6a3cbc6423e8ea05c0ae649e11a2f4059a84b3faa6cde0/","source_url":"https://arxiv.org/abs/2601.20861","text_available":true},{"id":"arxiv:2609.16454","title":"Fine-Tuning Fixes Mode Collapse and Over-Dispersion in LLMs","date":"2026-09-15","section":"Post-training","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/96b9ce3c41bf5be0185a59605d2cade1d5dcdc9432040a1195506fac45cb96fa/","source_url":"https://arxiv.org/abs/2609.16454","text_available":true},{"id":"arxiv:2610.02140","title":"Finetuning with Sampling: SFT Learns Better Than You Think","date":"2026-10-01","section":"Post-training","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/60f5edfae92f505bf86e8890135a44b4129825b0525b5ad8d242e8ba04182605/","source_url":"https://arxiv.org/abs/2610.02140","text_available":true},{"id":"arxiv:2509.25123","title":"From $f(x)$ and $g(x)$ to $f(g(x))$: LLMs Learn New Skills in RL by Composing Old Ones","date":"2025-09-29","section":"Post-training","quality":-0.057993,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7fdfe7abe26ee324fe7abad3ce802148163129a7c157d77bdec859c990868e03/","source_url":"https://arxiv.org/abs/2509.25123","text_available":true},{"id":"arxiv:2605.22074","title":"From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning","date":"2026-05-21","section":"Post-training","quality":0.202613,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/e51c03db071e17a4d5403433f1037e79c3cda795e0b20f6a490df35167411dd0/","source_url":"https://arxiv.org/abs/2605.22074","text_available":true},{"id":"arxiv:2507.19457","title":"GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning","date":"2025-07-25","section":"Post-training","quality":0.450458,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/c1ef72d96fa3292590a6a3d112ab509840081b59aaa771cf42b8b0e8ef62ba08/","source_url":"https://arxiv.org/abs/2507.19457","text_available":true},{"id":"arxiv:2602.18037","title":"Gradient Regularization Mitigates Reward Hacking in Reinforcement Learning from Human Feedback and Verifiable Rewards","date":"2026-02-20","section":"Post-training","quality":0.140712,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/75161700a0f3e03055116a289c82e9f9fa7d8fee7a6f802e116216eac3a70e31/","source_url":"https://arxiv.org/abs/2602.18037","text_available":true},{"id":"arxiv:2507.18071","title":"Group Sequence Policy Optimization","date":"2025-07-24","section":"Post-training","quality":0.050742,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/69430db904efc7978803673806cc9338004ffd8cf92199f3349dfbfa8a44e440/","source_url":"https://arxiv.org/abs/2507.18071","text_available":true},{"id":"arxiv:2505.10978","title":"Group-in-Group Policy Optimization for LLM Agent Training","date":"2025-05-16","section":"Post-training","quality":0.31133,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/2fa466eef15ebc0489217004e5ebb80efd37db9675a603331ccc3b7da055d147/","source_url":"https://arxiv.org/abs/2505.10978","text_available":true},{"id":"url:cameronrwolfe.substack.com/p/grpo-tricks","title":"GRPO++: Tricks for Making RL Actually Work (Cameron Wolfe)","date":"2026-01-05","section":"Post-training","quality":null,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/d170bc3da5b8e6a055fd511719fcc3764635236bac638947bd0a018412245eb5/","source_url":"https://cameronrwolfe.substack.com/p/grpo-tricks","text_available":false},{"id":"arxiv:2604.20659","title":"GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision for Effective Reasoning","date":"2026-04-22","section":"Post-training","quality":-0.004395,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/42caf0f9c992f5df8bac4d0f7e45c3a5bf57879661f41febd01a047694d52628/","source_url":"https://arxiv.org/abs/2604.20659","text_available":true},{"id":"arxiv:2602.09000","title":"iGRPO: Self-Feedback-Driven LLM Reasoning","date":"2026-02-09","section":"Post-training","quality":-0.04492,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/1cd6a66823fe24e8e2d83e817177d3a1e8cda1d7d5b59a4397f8669bee6306ae/","source_url":"https://arxiv.org/abs/2602.09000","text_available":true},{"id":"arxiv:2510.00977","title":"It Takes Two: Your GRPO Is Secretly DPO","date":"2025-10-01","section":"Post-training","quality":-0.095378,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/575241376f2282c1929931f107a0ce60fd25f0a5388ecfea6b9abbbed2f03ca6/","source_url":"https://arxiv.org/abs/2510.00977","text_available":true},{"id":"arxiv:2508.07629","title":"Klear-Reasoner: Advancing Reasoning Capability via Gradient-Preserving Clipping Policy Optimization","date":"2025-08-11","section":"Post-training","quality":-0.228184,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/a232533e4e78ad28d2ffe9fafe799b66cf626b18ad523d0088d9cc12f1ba1b76/","source_url":"https://arxiv.org/abs/2508.07629","text_available":true}],"next":"https://vinnik-dmitry07.github.io/favorite-papers/agents/catalog-12-2.html"}