LingBot-VA 2.0 Real-World Robot Performance Gains
rohanpaul_ai · x · 2026-07-14
This thread details the empirical results and system design of LingBot-VA 2.0:
- On RoboTwin 2.0, the authors report an average score of 93.6, with 93.8 for clean and 93.4 for randomized settings.
- In their custom real-world test set, a single generalist policy trained with 20 demonstrations per task achieved success rates of 77% / 67% / 82% / 64% across 4 tasks, outperforming the compared π0.5 (58% / 42% / 64% / 45%).
- Ultimately, the system emphasizes learning from few-shot demonstrations and transferring across robot morphologies.
Related event: LingBot-VA 2.0: A Control-Native Foundation Model for Robotics(9 posts)→
More from Embodied
- 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 hits 80.8% on RoboTwin tasks, beating strongest baseline by 16 points — trevordarrell · 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