Evidence-Augmented Policy Optimization with Reward Co-Evolution for Long-Context Reasoning
This addresses the challenge of improving evidence retrieval in long-context reasoning for LLM applications, representing a novel method for a known bottleneck.
The paper tackled the problem of sparse outcome rewards in reinforcement learning for long-context reasoning, which fails to penalize ungrounded guesses and leaves evidence retrieval unsupervised, and the result was that EAPO significantly enhanced performance across eight benchmarks compared to state-of-the-art baselines.
While Reinforcement Learning (RL) has advanced LLM reasoning, applying it to long-context scenarios is hindered by sparsity of outcome rewards. This limitation fails to penalize ungrounded "lucky guesses," leaving the critical process of needle-in-a-haystack evidence retrieval largely unsupervised. To address this, we propose EAPO (Evidence-Augmented Policy Optimization). We first establish the Evidence-Augmented Reasoning paradigm, validating via Tree-Structured Evidence Sampling that precise evidence extraction is the decisive bottleneck for long-context reasoning. Guided by this insight, EAPO introduces a specialized RL algorithm where a reward model computes a Group-Relative Evidence Reward, providing dense process supervision to explicitly improve evidence quality. To sustain accurate supervision throughout training, we further incorporate an Adaptive Reward-Policy Co-Evolution mechanism. This mechanism iteratively refines the reward model using outcome-consistent rollouts, sharpening its discriminative capability to ensure precise process guidance. Comprehensive evaluations across eight benchmarks demonstrate that EAPO significantly enhances long-context reasoning performance compared to SOTA baselines.