Humanoid robots just found their “pretraining” moment. 🤯!
This is a AI post classified by Jev as AI research (a model release), kept by the AI Radar because it carries real work, not commentary.
Humanoid robots just found their “pretraining” moment. 🤯! Figure AI just showed something I think could become a major turning point for physical AI.👀 Helix 2.5 was pretrained on Index, Figure’s large-scale human behavior dataset, then tested on 3 household tasks across 30 unseen homes without collecting data or adapting the model to those homes. The result: 56% end-to-end success, compared with just 8% for models trained without Index.👀 But the most interesting part isn’t the 56%. Figure found that increasing human pretraining data by 8× produced a predictable improvement in robot-action
Posted by SciTech Era (22.6k followers) 1 days ago · 128 likes · 11.6k views · view the original post on X. Kept by the AI Radar as AI research.
More AI work like this
- SoupFold from KAIST: no single co-folding model wins everywhere, so learn simple… — @biogerontology
- We built AI that actually knows aging biology. Longevity-LLM (9B params) beats OpenAI… — @biogerontology
- We need more examples like this in the open-source RL ecosystem — @adithya_s_k
- "Verbalizing Subliminal Learning Effects Using Text Optimization" — @askalphaxiv
- “What Does Privileged Information Add to On-Policy Self-Distillation?” — @askalphaxiv
- “Score Centering Stabilizes Off-Policy Reinforcement Learning” — @askalphaxiv
- Banger paper from Google Cloud AI Research. — @dair_ai
- If you liked the MiMo-V2.6 livestreamed RL run, don’t forget to check out Marin’s… — @adi_baradwaj
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 27.2k posts from 4.8k X accounts over the last 14 days, 1.2k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-19 18:50 UTC. Full method.