Papers › Task adaptation of Vision-Language-Action model: 1st Place Solution for the 2025...

Task adaptation of Vision-Language-Action model: 1st Place Solution for the 2025 BEHAVIOR Challenge

7 Dec 2025NeurIPSarXiv:2512.06951added by Syntology

Ilia Larchenko, Gleb Zarin, Akash Karnatak

Title, abstract, authors and date from arXiv's metadata (CC0); this paper is not in the Papers with Code archive (frozen 2025-07-28).

We present a vision-action policy that won 1st place in the 2025 BEHAVIOR Challenge - a large-scale benchmark featuring 50 diverse long-horizon household tasks in photo-realistic simulation, requiring bimanual manipulation, navigation, and context-aware decision making. Building on the Pi0.5 architecture, we introduce several innovations. Our primary contribution is correlated noise for flow matching, which improves training efficiency and enables correlation-aware inpainting for smooth action sequences. We also apply learnable mixed-layer attention and System 2 stage tracking for ambiguity resolution. Training employs multi-sample flow matching to reduce variance, while inference uses action compression and challenge-specific correction rules. Our approach achieves 26% q-score across all 50 tasks on both public and private leaderboards.

PaperPDF

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

IliaLarchenko/behavior-1k-s found in paper text by Syntologynot reachable when probed 2026-09-16 — repositories for recent papers often appear after camera-ready report

Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Syntology holds the repository link but has not harvested or run code from it.

Results from the paper

The Papers with Code archive ends with its 2025-07-28 snapshot. This paper's arXiv identifier, 2512.06951, was issued in December 2025, after that date, so the archive has no leaderboard rows for it.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections