A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Guanghao Ye, Khiem Pham, Xinzhi Zhang, et al.
Proposes RLSP, a post-training framework that decouples exploration and correctness signals during PPO to enable emergent reasoning behaviors in LLMs.
David Silver, Thomas Hubert, Julian Schrittwieser, et al.
AlphaZero is a single reinforcement learning algorithm that, starting from random play and given only game rules, achieves superhuman performance in chess, shogi, and Go through self-play.
Akarsh Kumar, Ryan Bahlous-Boldi, Prafull Sharma, et al.
This paper introduces Digital Red Queen (DRQ), a simple self-play algorithm that uses LLMs to evolve assembly programs in Core War, demonstrating convergent evolution toward general strategies and suggesting minimal self-play for adversarial domains.
Unknown
Self-Play Preference Optimization (SPO) is a minimalist reinforcement learning from human feedback algorithm that avoids training a separate reward model by using self-play.
Unknown
This paper provides a sharp theoretical analysis of model-based reinforcement learning with self-play, establishing sample complexity bounds and showing that model-based algorithms can achieve near-optimal performance in competitive environments.
Unknown
Spell introduces a self-play reinforcement learning framework that evolves long-context language models by generating and learning from progressively longer reasoning tasks.
Siyuan Huang, Pengyu Cheng, Haotian Liu, et al.
Skill Self-Play (Skill-SP) co-evolves a proposer, solver, and skill controller via RL to reconcile structured verification with open-ended task diversity for LLM self-evolution.