Hindsight Experience Replay

Latest papers

9 papers2 papers in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.

GTRL: Grounding Divide-and-Conquer Value Learning with Temporal Differences
Sep 27, 2026Abdul Monaf Chowdhury, MD Sameer Iqbal Chowdhury, Shifat E Arman +1Reinforcement Learning Temporal-Difference Learning
Department of Robotics and Mechatronics Engineering, University of Dhaka, Bangladesh · Department of Computer Science, Texas State University, USA · Department of Computer Science, University of Oxford, UK
On Preference Coverage Collapse from Hindsight Relabeling in Multi-Objective Reinforcement Learning
Sep 22, 2026Baptiste Bonin, Caro Strickland, Audrey DurandMulti-Objective Reinforcement Learning Experience Replay
Mila (Quebec AI Institute) · Université Laval · CIFAR AI Chair
Iron: Intent-Aligned and Retrospective Dual Learning Framework for Enhancing Generalist Virtual Agents
Aug 28, 2026Jiahe Ying, Wendong Bu, Kaihang Pan +6Hindsight Experience Replay LLM Agent Training
Fudan University · Zhejiang University
Learning from Hindsight for VLA Reinforcement Learning
Jul 10, 2026Iris Xu, Sunshine Jiang, John Marangola +2Vision-Language-Action Models RL for VLMs
Massachusetts Institute of Technology · MIT-IBM Computing Research Lab
Goal Sets, Not Goal States: Queryable Robot Goals through Goal-Set Hindsight Relabeling
Jun 8, 2026Carlos Vélez García, Miguel Cazorla, Jorge PomaresOffline RL Goal-Conditioned RL
INESCOP · University of Alicante
AgentHER: Hindsight Experience Replay for LLM Agent Trajectory Relabeling
Mar 22, 2026Liang DingHindsight Experience Replay LLM Agent Training
The University of Sydney
Experience is the Best Teacher: Motivating Effective Exploration in Reinforcement Learning for LLMs
Mar 20, 2026Wenjian Zhang, Kongcheng Zhang, Jiaxin Qi +2Hindsight Experience Replay RL for Language Model Reasoning
Dalian University of Technology · Zhejiang University · Chinese Academy of Sciences
CONTHER: Context-Aware Reinforcement Learning for Robotic Manipulation with Sparse Rewards
Mar 20, 2025Maria Makarova, Qian Liu, Dzmitry TsetserukouReinforcement Learning Experience Replay
Intelligent Space Robotics Laboratory, Center for Digital Engineering, Skolkovo Institute of Science and Technology (Skoltech), 121205 Moscow, Russia · Department of Computer Science and Technology, Dalian University of Technology, China
ETHER: Aligning Emergent Communication for Hindsight Experience Replay
Jul 28, 2023Kevin Yandoka Denamganaï, Daniel Hernandez, Ozan Vardal +2Reinforcement Learning Experience Replay
Department of Computer Science University of York York, UK

Sep 27, 2026J/K move · Enter open · S save

GTRL: Grounding Divide-and-Conquer Value Learning with Temporal Differences

Abdul Monaf Chowdhury, MD Sameer Iqbal Chowdhury, Shifat E Arman, Md Mehedi Hasan

Department of Robotics and Mechatronics Engineering, University of Dhaka, Bangladesh · Department of Computer Science, Texas State University, USA · Department of Computer Science, University of Oxford, UK

Reinforcement Learning Temporal-Difference Learning Offline RL Goal-Conditioned RL Hindsight Experience Replay

Open paper AskPDF arXiv

In offline goal-conditioned reinforcement learning (GCRL), divide-and-conquer scales to long horizons by joining two shorter segments at a subgoal. However, under stochastic dynamics, the base case of this rule values the luckiest trajectories through the data. The subgoal must also lie on a shared trajectory, so a state-goal pair that no trajectory connects gets no value update at all. To address both, we present Grounded Transitive RL (GTRL), an offline GCRL value learning algorithm that grounds the divide-and-conquer update with a one-step TD target. Over a single step, TD is correct, as its target averages over the successors and needs no subgoal. GTRL adds this target to the composition rather than replacing it, so every pair receives an update, and the composition still carries the long horizon. GTRL also corrects the bias from hindsight relabeling by reweighting each goal against how reachable it was from other successors. We evaluate our algorithm on nineteen OGBench tasks spanning stochastic, deterministic, and stitching environments, where it achieves the highest average success rate. Code will be released soon.