Showing 2026Show all
2 papers · 1 filter
cs.LG2026
When Does External Guidance Help LLM Reasoning? A Bias-Variance Theory of Guidance-Augmented GRPO
Sofia Torres, Gabriel Almeida, Carter Adams +1
Reinforcement learning with verifiable rewards (RLVR) has become the dominant paradigm for eliciting multi-step reasoning in large language models, and a recent wave of methods (LU…
cs.LG2026
Rethinking Reinforcement Fine-Tuning in LVLM: Convergence, Reward Decomposition, and Generalization
Carter Adams, Rafael Oliveira, Gabriel Almeida +1
Reinforcement fine-tuning with verifiable rewards (RLVR) has emerged as a powerful paradigm for equipping large vision-language models (LVLMs) with agentic capabilities such as too…