Image de couverture de livre

How Reasoning Evolves from Post-Training Data: An Empirical Study Using Chess

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Dionisopoulos, Lucas, Majamaki, Nicklas, Ammanabrolu, Prithviraj
Format:	Preprint
Publié:	2026
Sujets:	Machine Learning Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Beyond Needle(s) in the Embodied Haystack: Environment, Architecture, and Training Considerations for Long Context Reasoning
par: Kim, Bosung, et autres
Publié: (2025)

A Practitioner's Guide to Multi-turn Agentic Reinforcement Learning
par: Wang, Ruiyi, et autres
Publié: (2025)

Preference-Based Learning in Audio Applications: A Systematic Analysis
par: Broukhim, Aaron, et autres
Publié: (2025)

Simultaneous Multi-objective Alignment Across Verifiable and Non-verifiable Rewards
par: Shen, Yiran, et autres
Publié: (2025)

Introspective X Training: Feedback Conditioning Improves Scaling Across all LLM Training Stages
par: Cui, Brandon, et autres
Publié: (2026)

VAM: Verbalized Action Masking for Controllable Exploration in RL Post-Training -- A Chess Case Study
par: Zhang, Zhicheng, et autres
Publié: (2026)

LaDiR: Latent Diffusion Enhances LLMs for Text Reasoning
par: Kang, Haoqiang, et autres
Publié: (2025)

Scaling Behaviors of LLM Reinforcement Learning Post-Training: An Empirical Study in Mathematical Reasoning
par: Tan, Zelin, et autres
Publié: (2025)

Can Large Language Models Develop Strategic Reasoning? Post-training Insights from Learning Chess
par: Hwang, Dongyoon, et autres
Publié: (2025)

ChessArena: A Chess Testbed for Evaluating Strategic Reasoning Capabilities of Large Language Models
par: Liu, Jincheng, et autres
Publié: (2025)

Zipping the Thought: When and How Compressed Reasoning Data Works in LLM Post-Training
par: Matsutani, Kohsei, et autres
Publié: (2026)

Behavior Cue Reasoning: Monitorable Reasoning Improves Efficiency and Safety through Oversight
par: Cui, Christopher Z., et autres
Publié: (2026)

ChessQA: Evaluating Large Language Models for Chess Understanding
par: Wen, Qianfeng, et autres
Publié: (2025)

Data-Efficient Training by Evolved Sampling
par: Cheng, Ziheng, et autres
Publié: (2025)

Front-Loading Reasoning: The Synergy between Pretraining and Post-Training Data
par: Akter, Syeda Nahida, et autres
Publié: (2025)

Implicit Search via Discrete Diffusion: A Study on Chess
par: Ye, Jiacheng, et autres
Publié: (2025)

How Instruction and Reasoning Data shape Post-Training: Data Quality through the Lens of Layer-wise Gradients
par: Li, Ming, et autres
Publié: (2025)

Generating Creative Chess Puzzles
par: Feng, Xidong, et autres
Publié: (2025)

Liquid Reasoning Transformers: A Sudoku-Based Prototype for Chess-Scale Algorithmic Tasks
par: Sahni, Shivansh, et autres
Publié: (2025)

Amortized Planning with Large-Scale Transformers: A Case Study on Chess
par: Ruoss, Anian, et autres
Publié: (2024)

Theoretical Perspectives on Data Quality and Synergistic Effects in Pre- and Post-Training Reasoning Models
par: Javanmard, Adel, et autres
Publié: (2026)

Dr. Post-Training: A Data Regularization Perspective on LLM Post-Training
par: Hu, Pingbang, et autres
Publié: (2026)

Complete Chess Games Enable LLM Become A Chess Master
par: Zhang, Yinqi, et autres
Publié: (2025)

Human-aligned Chess with a Bit of Search
par: Zhang, Yiming, et autres
Publié: (2024)

Enhancing Chess Reinforcement Learning with Graph Representation
par: Rigaux, Tomas, et autres
Publié: (2024)

An Empirical Study of Data Ability Boundary in LLMs' Math Reasoning
par: Chen, Zui, et autres
Publié: (2024)

Layer Importance for Mathematical Reasoning is Forged in Pre-Training and Invariant after Post-Training
par: Nepal, Aadim, et autres
Publié: (2025)

Towards Piece-by-Piece Explanations for Chess Positions with SHAP
par: Spinnato, Francesco
Publié: (2025)

Iterative Inference in a Chess-Playing Neural Network
par: Sandmann, Elias, et autres
Publié: (2025)

Diversifying AI: Towards Creative Chess with AlphaZero
par: Zahavy, Tom, et autres
Publié: (2023)

Mastering Chinese Chess AI (Xiangqi) Without Search
par: Chen, Yu, et autres
Publié: (2024)

Self-Evolving Curriculum for LLM Reasoning
par: Chen, Xiaoyin, et autres
Publié: (2025)

Reasoning Language Model Inference Serving Unveiled: An Empirical Study
par: Li, Qi, et autres
Publié: (2025)

Oracle-Guided Soft Shielding for Safe Move Prediction in Chess
par: Rajendran, Prajit T, et autres
Publié: (2026)

Mixture of Masters: Sparse Chess Language Models with Player Routing
par: Frisoni, Giacomo, et autres
Publié: (2026)

Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training
par: Xu, Yuanda, et autres
Publié: (2026)

Evaluating In Silico Creativity: An Expert Review of AI Chess Compositions
par: Veeriah, Vivek, et autres
Publié: (2025)

Learning to Reason Efficiently with A* Post-Training
par: Opedal, Andreas, et autres
Publié: (2026)

Post-Training as Reweighting: A Stochastic View of Reasoning Trajectories in Language Models
par: Bu, Dake, et autres
Publié: (2025)

R-Zero: Self-Evolving Reasoning LLM from Zero Data
par: Huang, Chengsong, et autres
Publié: (2025)