MEMOBench: A Process Level Memory Benchmark for Robotic Manipulation
Quick summary
arXiv:2609.07047v1 Announce Type: cross Abstract: Robotic manipulation often requires acting on information that is no longer visible, yet Vision-Language-Action policies are usually evaluated when the current observation largely determines the next action. Existing robotic memory benchmarks expose this gap, but they still rely mainly on final task success and therefore conflate forgetting with manipulation failure. We present \textbf{MEMOBench}, a benchmark for process level memory evaluation in robotic manipulation. MEMOBench includes 30 history dependent tasks, 1{,}500 expert demonstrations
Key takeaways
- arXiv:2609.07047v1 Announce Type: cross Abstract: Robotic manipulation often requires acting on information that is no longer visible, yet Vision-Language-Action policies are usually evaluated when the current observation largely determines the next action.
- Existing robotic memory benchmarks expose this gap, but they still rely mainly on final task success and therefore conflate forgetting with manipulation failure.
- We present \textbf{MEMOBench}, a benchmark for process level memory evaluation in robotic manipulation.
Why it matters
“MEMOBench: A Process Level Memory Benchmark for Robotic Manipulation” highlights the need for repeatable measurement rather than a single impressive demonstration. Independent validation across datasets and clearly stated limitations determine whether a result can guide product decisions.
Member comments