GPT-6-Astra as Robot Policy for RoboMME
Overview
A research paper presents the use of a frozen frontier vision-language model, GPT-6-Astra, as a robot policy for memory-dependent manipulation tasks on the RoboMME benchmark. The model achieves a success rate of 92.1% on 800 test episodes, outperforming human performance and existing systems, with failures concentrated in two tasks.
From arXiv Robotics — research abstracts
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · tasks matched or exceeded: 13 other · Basis not reported
“est episodes it succeeds in 92.1%, above the previously reported human score of 90.5%, which was obtained with an oracle skill interface, and far ahead of the best leaderboard entry (60.8%) and of a hierarchical system that pairs the same model with a RoboMME-trained policy (79.1%). It matches or exceeds the human score on 13 of the 16 tasks. Its 63 failures concentrate in two tasks and are mostly premature completio”
Exact source · revision 2Source owner not reported
Reported quantity · failures: 63 other · Basis not reported
“est episodes it succeeds in 92.1%, above the previously reported human score of 90.5%, which was obtained with an oracle skill interface, and far ahead of the best leaderboard entry (60.8%) and of a hierarchical system that pairs the same model with a RoboMME-trained policy (79.1%). It matches or exceeds the human score on 13 of the 16 tasks. Its 63 failures concentrate in two tasks and are mostly premature completio”
Exact source · revision 2Source owner not reported
Reported quantity · tasks with failures: 2 other · Basis not reported
“est episodes it succeeds in 92.1%, above the previously reported human score of 90.5%, which was obtained with an oracle skill interface, and far ahead of the best leaderboard entry (60.8%) and of a hierarchical system that pairs the same model with a RoboMME-trained policy (79.1%). It matches or exceeds the human score on 13 of the 16 tasks. Its 63 failures concentrate in two tasks and are mostly premature completio”
Exact source · revision 2Source owner not reported
Evidence dependency changes (3)
- 2026-10-30T23:06:09.945Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T23:05:54.286Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-19T19:27:45.810Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalCan a Frontier VLM Serve as a Robot Policy for Memory-Dependent Manipulation?
A research paper presents the use of a frozen frontier vision-language model, GPT-6-Astra, as a robot policy for memory-dependent manipulation tasks on the RoboMME benchmark. The model achieves a success rate of 92.1% on 800 test episodes, outperforming human performance and existing systems, with failures concentrated in two tasks.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).