MemBodied: Recurrent Associative Memory for Vision-Language-Action Models
MemBodied gives vision-language-action policies fixed episodic memory without expanding context.
MemBodied adds a fixed-size episodic memory to vision-language-action policies so history-dependent manipulation does not require storing every past observation. It combines an associative state of interactions across policy calls with an episode anchor that keeps a compact view of the initial scene. On five memory-dependent RMBench tasks it achieves 7.81 times the mean success of a stateless policy and 2.98 times vanilla recurrent memory, beating the strongest memory baseline by 1.3 times with 10 times fewer added parameters. On fully observable LIBERO-Long it scores 90.6%, a 5.4-point gain over the stateless pi0 policy.
- Fixed memory avoids ever-growing context and inference latency
- 7.81 times stateless success on five RMBench memory tasks
- 1.3 times the strongest baseline with 10 times fewer parameters
- LIBERO-Long reaches 90.6%, 5.4 points above stateless pi0
Full article197 words · extracted from huggingface.co · click to collapse
Vision-Language-Action models provide a strong foundation for general-purpose robot control, yet a vast majority of policies do not preserve and leverage episode-level information beyond the current observation. This limitation is consequential in history-dependent manipulation tasks that depend on information available only in past observations. Retaining past observations in context can aid in recovering this information, but at the significant cost of ever-growing, bloated context and inference latency. We thus introduce MemBodied, a fixed-size episodic memory with two complementary components: an associative state that records interactions across policy calls and an episode anchor that preserves a compact representation of the initial scene as a reference. At each policy call, the model conditions action generation on the current input and the memory components, rather than directly using past observations. Across five evaluated RMBench tasks requiring memory, MemBodied achieves 7.81times the mean success rate of a stateless policy and 2.98times of vanilla recurrent memory, while outperforming the strongest memory-augmented baseline by 1.3times with 10times fewer added parameters. On the fully observable LIBERO-Long suite, it reached 90.6%, a 5.4% improvement over the stateless π_0 policy. These findings support MemBodied as a practical alternative to expanding the policy context for history-dependent manipulation.
Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/papers/2609.28256