DRAM: 로봇 조작 정책을 위한 Delta-rule Recurrent Associative Memory

[Submitted on 26 Sep 2026 (v1), last revised 29 Sep 2026 (this version, v2)]

View PDF HTML (experimental)

Abstract:Robotic manipulation is inherently history-dependent, yet most pretrained robotic policies condition on only the current observation or a short temporal window. Equipping such policies with long-term memory remains challenging: existing approaches either feed the backbone multi-frame observation windows, which substantially increase inference cost, or rely on pre-defined semantic features, which limit task generality and may also require the retraining of the backbone to adapt to the memory. We introduce DRAM (Delta-rule Recurrent Associative Memory), a plug-and-play memory module that can be attached to a wide range of pretrained robotic policies, endowing them with long-horizon memory without architectural modification or backbone retraining, requiring only task-specific post-training of the memory module and action expert. DRAM maintains a fixed-size associative memory using gated delta-rule linear attention, with a modified update that incorporates all tokens within each frame in parallel. An architecture-agnostic readout integrates historical context into action prediction across different policy architectures. Experiments show that DRAM consistently improves frozen pretrained policies over short-context baselines and alternative compact memory designs, validating its effectiveness as a fixed-size, post-hoc memory module trained with the backbone frozen.

Submission history

From: Yixiang Shan [view email]
[v1] Sat, 26 Sep 2026 10:38:49 UTC (12,797 KB)
[v2] Tue, 29 Sep 2026 03:00:59 UTC (6,826 KB)