Xinran Chen
Publications
TurnSight: Turn-Level Hindsight Self-Distillation for Tool-Integrated Reasoning
Tool-Integrated Reasoning (TIR) enables LLMs to solve complex tasks through iterative tool interactions. However, existing reinforcement learning methods often rely on trajectory-level supervision, limiting fine-grained credit assignment in long-horizon TIR scenarios. On-policy self-distillation offers denser signals through teacher branches with privileged context, but existing approaches typically derive such context from ground-truth answers or retrieved skills, which may not reflect the states actually visited by the agent. Moreover, token-level supervision fails to capture the turn-level structure of tool interactions. To address this, we propose TurnSight, a turn-level hindsight self-distillation framework that derives supervision directly from execution-conditioned hindsight. It then constructs multiple hindsight views with different lookahead horizons and selects reliable supervision through cross-horizon directional agreement. Finally, the selected hindsight signal is normalized across sibling rollouts and used to adaptively modulate RL advantages while preserving their original optimization direction. Extensive experiments on three benchmarks demonstrate the effectiveness of TurnSight. Our codes are available at https://github.com/quchangle1/TurnSight.
STAMP: Provenance-Guided Credit Assignment for Deep Search Agents
Reinforcement learning for deep-search agents has largely focused on trajectory-level scoring -- outcome correctness, citation-aware rewards, and evidence coverage. Yet the actions that expose supporting documents receive no targeted credit, a gap we call the reward-credit mismatch. We propose STAMP, in which a reference-based verifier judges whether each cited document supports an entity or relation in a training-time evidence graph, and first-exposure attribution traces each supported citation back to the action that first surfaced it. This step credit is injected through sign-preserving advantage modulation, which redistributes advantage across steps without changing the trajectory-level reward or the relative ranking of trajectories within each group. On BrowseComp, BrowseComp-ZH, and xbench-DS, STAMP improves the GRPO baseline by +2.0/+5.5/+3.0 points under matched SFT initialization, training data, and search tools, and composes with both outcome-only and citation-rubric base rewards. Component ablations confirm that the provenance-based credit signal and the sign-preserving advantage modulation each contribute to the gains.
Not All Preferences Are Created Equal: Stability-Aware and Gradient-Efficient Alignment for Reasoning Models
Preference-based alignment is pivotal for training large reasoning models; however, standard methods like Direct Preference Optimization (DPO) typically treat all preference pairs uniformly, overlooking the evolving utility of training instances. This static approach often leads to inefficient or unstable optimization, as it wastes computation on trivial pairs with negligible gradients and suffers from noise induced by samples near uncertain decision boundaries. Facing these challenges, we propose SAGE (Stability-Aware Gradient Efficiency), a dynamic framework designed to enhance alignment reliability by maximizing the Signal-to-Noise Ratio of policy updates. Concretely, SAGE integrates a coarse-grained curriculum mechanism that refreshes candidate pools based on model competence with a fine-grained, stability-aware scoring function that prioritizes informative, confident errors while filtering out unstable samples. Experiments on multiple mathematical reasoning benchmarks demonstrate that SAGE significantly accelerates convergence and outperforms static baselines, highlighting the critical role of policy-aware, stability-conscious data selection in reasoning alignment.
Tracking Drift: Variation-Aware Entropy Scheduling for Non-Stationary Reinforcement Learning
Real-world reinforcement learning often faces environment drift, but most existing methods rely on static entropy coefficients/target entropy, causing over-exploration during stable periods and under-exploration after drift (thus slow recovery), and leaving unanswered the principled question of how exploration intensity should scale with drift magnitude. We prove that entropy scheduling under non-stationarity can be reduced to a one-dimensional, round-by-round trade-off, faster tracking of the optimal solution after drift vs. avoiding gratuitous randomness when the environment is stable, so exploration strength can be driven by measurable online drift signals. Building on this, we propose AES (Adaptive Entropy Scheduling), which adaptively adjusts the entropy coefficient/temperature online using observable drift proxies during training, requiring almost no structural changes and incurring minimal overhead. Across 4 algorithm variants, 12 tasks, and 4 drift modes, AES significantly reduces the fraction of performance degradation caused by drift and accelerates recovery after abrupt changes.