AW-0 hits 80.8% on RoboTwin tasks, beating strongest baseline by 16 points
trevordarrell · x · 2026-10-27
Part 7/8 of the AW-0 thread: evaluation results across 50 RoboTwin 2.0 tasks and 8 real-world tasks using YAM parallel grippers and Tianji-Sharpa dexterous hands. AW-0 reaches 80.8% success on RoboTwin's clean setting vs 64.8% for the strongest baseline, and scores 71.1 average rubric on real-world tasks vs 46.0. Gains span instruction-specific object selection, counting, tool use, and coordinated manipulation.
More from Embodied
- UniAction: unified action conditioning lets one world model handle mixed action types — ZeYanjie · 2026-10-27
- StackChan robot performs an AI-generated skit, orchestrated by Hermes Agent via JSON script — alexcovo_eth · 2026-10-27
- DONIC hits 86.79% on bimanual dexterous tracking, crushing prior SOTA's 9.43% — liuziwei7 · 2026-10-27
- Bedrock Robotics raises $600M as its fully autonomous excavators do paid work on live job sites — Sethwinterroth · 2026-10-27
- AW-0's code stream fixes world model prediction: picks mango where baseline grabs apple — trevordarrell · 2026-10-27
- AW-0 runs at 230ms on a single RTX 5090, 2.4x faster than autoregressive code planning — trevordarrell · 2026-10-27