Its task suites separate counting, object permanence, reference and imitation demands. These distinctions explain failures that a single headline average would hide.
Yinpei Dai and coauthors · Publication date not disclosed · Source accessed 2026-09-28
Reading results
The paper studies memory-augmented variants built on π0.5. Those modified policies are not the unmodified public base checkpoint, and their results must preserve memory design and training configuration.
RoboMME · Publication date not disclosed · Source accessed 2026-09-28
Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.
Reported benchmark results
Context before scores. Results from different benchmarks are not directly comparable. A simulation result does not establish real-world reliability, safety or commercial availability.
No result meets our complete revision and methodology requirements for this record yet. Inspect the original benchmark documentation before comparing published scores.
What this evidence does not establish
The default released environment focuses on imitation learning; alternative action interfaces can change the evaluated problem.
Related reading is an editorial crosslink. Sourced connections describe relationships reported in the cited material. A link to a versioned profile does not establish compatibility with that version unless the connection note explicitly identifies it.
Record history & verified changes
A research review records when we checked a source. It does not mark a product launch or a new deployment.
Initial reviewed record. No subsequent field change has been recorded.