One Policy for 100 Household Tasks: Our Solution for the 2026 BEHAVIOR Challenge
Overview
This paper presents a solution for the 2026 BEHAVIOR Challenge, utilizing a single vision-language-action (VLA) model to perform 100 long-horizon household tasks in the OmniGibson simulator. The team achieved 9th place on the final leaderboard. Key improvements included extended training over 200k steps, optimization of training parameters, and the integration of a GPT-6 Astra coding agent for 14 specific tasks where it outperformed the learned policy.
Generated from attributed reports · Updated 2 hours ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Environment: simulation
“arXiv:2611.14432v1 Announce Type: new Abstract: We describe our solution for the 2026 BEHAVIOR Challenge: 100 long-horizon household tasks in the OmniGibson simulator. We placed 9th on the final leaderboard, the only independent team in the top 10. The core of the solution is a single vision-language-action (VLA) model, one checkpoint for all 100 tasks, built on our 1st-place 2025 solution, which is based on $\pi_{0.”
Exact source · revision 1Source owner not reported
Control: teleoperated
“demonstrations' skill annotations, task and stage embeddings of new tasks warm-started from similar old ones, and a 21-dimensional action space matching the teleoperation controls); variance-reduced multi-sample flow matching with a simple noise-vs-compute model that picks the number of action-expert samples per observation without ablations; and a 200k-step training run, about one pass over the 211M frames of the de”
Exact source · revision 1Source owner not reported
Reported quantity · training steps: 200000 other · Basis not reported
“demonstrations' skill annotations, task and stage embeddings of new tasks warm-started from similar old ones, and a 21-dimensional action space matching the teleoperation controls); variance-reduced multi-sample flow matching with a simple noise-vs-compute model that picks the number of action-expert samples per observation without ablations; and a 200k-step training run, about one pass over the 211M frames of the de”
Exact source · revision 1Source owner not reported
Reported quantity · frames: 211000000 other · Basis not reported
“demonstrations' skill annotations, task and stage embeddings of new tasks warm-started from similar old ones, and a 21-dimensional action space matching the teleoperation controls); variance-reduced multi-sample flow matching with a simple noise-vs-compute model that picks the number of action-expert samples per observation without ablations; and a 200k-step training run, about one pass over the 211M frames of the de”
Exact source · revision 1Source owner not reported
Reported quantity · tasks: 100 other · reported trials · Differing source assertions
“arXiv:2611.14432v1 Announce Type: new Abstract: We describe our solution for the 2026 BEHAVIOR Challenge: 100 long-horizon household tasks in the OmniGibson simulator. We placed 9th on the final leaderboard, the only independent team in the top 10. The core of the solution is a single vision-language-action (VLA) model, one checkpoint for all 100 tasks, built on our 1st-place 2025 solution, which is based on $\pi_{0.”
Exact source · revision 1Source owner not reported
Reported quantity · tasks: 50 other · reported trials · Differing source assertions
“5}$. On the 50 tasks shared with 2025, its Q rises from 0.260 to 0.371; across all 100 tasks it scores 0.392. Our main conclusion from 2025 was that the model was undertrained, so most of the 2026 work went into training the model for longer, optimizing some training parameters, making each training step slightly faster and reducing the gradient noise: a migration from 50 to 100 tasks (semantic stages built from the”
Exact source · revision 1Source owner not reported
Reported quantity · tasks: 14 other · reported trials · Differing source assertions
“monstrations. A cloud evaluation infrastructure keeps a shared record with provenance, episode diagnostics and restorable simulator states, and runs six simulators on one GPU together faster than real time. On top of the learned policy, 14 tasks are routed to a GPT-6 Astra coding agent that controls the robot through observation and motion tools, starting every episode from a frozen copy of its own practice session o”
Exact source · revision 1Source owner not reported
Reported quantity · Q score: 0.392 other · Basis not reported
“5}$. On the 50 tasks shared with 2025, its Q rises from 0.260 to 0.371; across all 100 tasks it scores 0.392. Our main conclusion from 2025 was that the model was undertrained, so most of the 2026 work went into training the model for longer, optimizing some training parameters, making each training step slightly faster and reducing the gradient noise: a migration from 50 to 100 tasks (semantic stages built from the”
Exact source · revision 1Source owner not reported
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalOne Policy for 100 Household Tasks: Our Solution for the 2026 BEHAVIOR Challenge
This paper presents a solution for the 2026 BEHAVIOR Challenge, utilizing a single vision-language-action (VLA) model to perform 100 long-horizon household tasks in the OmniGibson simulator. The team achieved 9th place on the final leaderboard. Key improvements included extended training over 200k steps, optimization of training parameters, and the integration of a GPT-6 Astra coding agent for 14 specific tasks where it outperformed the learned policy.
Event coverage history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).