← Back to all news
arXiv AI By Jiaxin Zhang, Xiangyu Peng, Qinglin Chen, Yu Li, Hiroaki Hayashi, Chien-Sheng Wu

Prospective Hindsight: Self-Calibrating Reinforcement Learning via Prediction-Reality Gaps

Read the original on arXiv AI →

Prospective Hindsight (PH) is a self‑calibrating training principle for reinforcement learning that augments any retrospective base method with a signal derived from the gap between an agent’s prospective prediction and its retrospective evaluation. By weighting gradients with a stop‑gradient surprise based on this prediction‑reality gap, PH highlights samples where the agent’s self‑model is most inaccurate and encourages the policy and predictor to co‑evolve, progressively focusing on remaining blind spots. Experiments on single‑turn verifiable tasks and a multi‑turn personal‑agent task show that PH improves both task performance and calibration across model scales, addressing overconfident failures in single‑turn settings and underconfident successes in multi‑turn settings.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv AI
Oct 7

Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight

The paper introduces Self‑Retrospection Distillation (SRD), a method that uses post‑hoc trajectories to supervise foresight predictions made before an agent acts. SRD distills information from completed rollouts into trajectory‑blind predictions, enabling the agent to anticipate useful actions and avoid pitfalls without requiring explicit foresight at inference time. Experiments on ten tool‑integrated reasoning and long‑horizon tasks show that SRD improves performance by up to 24.2 percentage points, especially when reward signals are scarce or uniform across rollouts.

By Haoxiang Zhang, Qinglin Chen, Hiroaki Hayashi, Zhuofeng Li, Siming Zhang, Jiaxin Zhang, Jixuan Chen, Fang Wu, Pan Lu, Silvio Savarese, Julian McAuley, Chien-Sheng Wu
agentsreinforcement-learningefficiency
More like this →
arXiv AI
Oct 1

Beyond Prediction: Steering VLM Agents with Retrospective World Modeling

The paper introduces Retrospective World Modeling, a new paradigm for vision‑language‑model (VLM) agents that allows them to reason backward by estimating which action most likely caused a state transition. It proposes the Self‑Consistency Reward (SCR), an intrinsic signal that measures how well a policy action aligns with this retrospective explanation, providing dense transition‑level feedback. Experiments demonstrate that incorporating SCR improves policy robustness and generalization compared to purely prospective world‑modeling approaches.

By Yongjiang Liu, Jie Zhang, Haoyue Zhang, Jingcai Guo, Deze Zeng, Song Guo
agentsreinforcement-learningmultimodal
More like this →
arXiv AI
Jul 15

Critic Experience Bank: Self-Evolving Step-Level Confidence Estimation for LLM Agents

arXiv:2607. 12397v1 Announce Type: new Abstract: LLM agents act in external environments where each action changes the state that later decisions condition on, and where a single wrong step can waste interaction budget or trigger irreversible side effects long before the final failure is observed.

By Yaopei Zeng, Congchao Wang, JianHang Chen, Nan Wang, Yurui Chang, Lu Lin
llmsagentsbenchmarks
More like this →
Hugging Face Trending Papers
Jul 14

Critic Experience Bank: Self-Evolving Step-Level Confidence Estimation for LLM Agents

LLM agents act in external environments where each action changes the state that later decisions condition on, and where a single wrong step can waste interaction budget or trigger irreversible side effects long before the final failure is observed. Reliable deployment therefore requires \emph{step-level confidence estimation}: a calibrated probability that each proposed action is productive, available \emph{before} the action is executed.

llmsagentsbenchmarks
More like this →