Records found: 7
SRL converts expert trajectories into per-step rewarded actions and lets models produce private reasoning spans before each action, giving dense learning signals that boost 7B open models on hard math and coding tasks
'RA3 formalizes mid-training as pruning plus horizon shortening and uses temporal action abstractions to accelerate RL post-training, boosting code generation benchmarks.'
K2 Think is a 32B open-weight reasoning system from MBZUAI that matches or outperforms much larger open models on math by combining long CoT fine-tuning, RL with verifiable rewards, planning, and wafer-scale speculative decoding.
Master-RM is a new reward model designed to fix vulnerabilities in LLM-based evaluators by reducing false positives caused by superficial cues, ensuring more reliable reinforcement learning outcomes.
‘Selective training on high-entropy tokens in LLMs improves reasoning performance and reduces computational costs, setting new benchmarks on AIME tests.’
Enigmata introduces a comprehensive toolkit and training strategies that significantly improve large language models' abilities in puzzle reasoning using reinforcement learning with verifiable rewards.
Researchers reveal that training large language models with just one example using 1-shot reinforcement learning significantly enhances their math reasoning abilities, matching results from large datasets.