MemHarness Argues Agent Memory Should Be Reconstructed, Not Replayed Verbatim

arXiv·medium signalXBlueskyLinkedInCopy link

Memory-augmented LLM agents typically inject retrieved experiences into context as static records regardless of whether they fit the current situation, which the authors argue causes negative transfer between abstract stored experience and concrete decision-time state. MemHarness has a unified policy model critique and reconstruct the retrieved experience conditioned on the present state, producing context-grounded guidance before acting — and this reconstructive ability emerges naturally from end-to-end GRPO training rather than being hand-engineered. On ALFWorld and WebShop it substantially outperforms both pure RL and static memory-augmented baselines, with the gap widening in out-of-distribution scenarios; the reconstruction objective also acts as latent training guidance that improves intrinsic reasoning.