Figure AI - Helix 2.5 Robot: Zero-Shot Home Generalization
Figure AI has introduced Helix 2.5, a neural network that enables humanoid robots to perform complex whole-body tasks in unseen environments without prior data collection. The system demonstrates zero-shot generalization across 30 real homes by leveraging pretraining on Figure's global-scale human behavior dataset called Index. This approach significantly improves success rates and reduces the need for task-specific data compared to previous models.
- ▪Helix 2.5 is the first system to demonstrate zero-shot whole-body generalization across 30 unseen homes for tasks like tidying, making beds, and folding towels.
- ▪The model was pretrained on Index, a global-scale dataset of human behavior, which increased zero-shot success rates from 9% to 56% compared to models trained from scratch.
- ▪A single foundation model was adapted to three distinct behaviors that require locomotion, rigid and deformable manipulation, and bimanual coordination.
- ▪Helix 2.5 used half as much task-specific data as previous Helix 02 behaviors while generalizing across 30 unseen homes.
- ▪Repeatedly doubling the Index pretraining data allowed researchers to forecast downstream robot-action prediction loss with high accuracy before training.
Hacker News (AI / LLM) files mainly under ai. We currently carry 5,388 of its stories.
Story provenance
Source · retrieval · rights · ranking — open for full record
inspect →
Story provenance
Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.
Record
| Original publisher | FigureAI |
| Canonical URL | https://www.figure.ai/news/helix-2-5-zero-shot-30-home-generalization |
| Publication time | Thu, 17 Sep 2026 19:37:23 +0000 |
| Retrieval time | 2026-09-17T19:43:45.074Z |
| Last seen | 2026-09-17T19:43:45.074Z |
| Headline source | Publisher (no WeSearch rewrite) |
| Excerpt source | publisher body |
| Excerpt method | First ~120 words (~800 chars) of extracted publisher body, fair-use limited. |
| Summary | WeSearch · cerebras-chat (WeSearch summarizer) |
| Summary source text | contentText |
| Citation coverage | Summary is a WeSearch-generated derivative; primary citation is the original publisher URL. |
| Cluster | jEDgxlTQig5h · 1 stories |
| Cluster logic | Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison. |
| Ranking reason | Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking. |
| Publisher visit | Yes — open original |
| Substitutes article? | No — link-out required for full text |
Rights status (four layers)
WeSearch handling by dimension
| Indexing | May the item be indexed (stored, ranked, made findable)? | Allowed |
| Snippet | May a short excerpt of the publisher's text be shown? | Allowed |
| AI summary | May WeSearch generate its own short summary of the article? | Limited |
| Retrieval / RAG | May the content be exposed for third-party retrieval-augmented generation? | Not asserted |
| Model training | May the content be used to train AI models? | Not asserted |
| Commercial reuse | May the content be reused commercially? | Not permitted |
Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.
Opening excerpt (first ~120 words) tap to expand
Helix 2.5: Zero-Shot 30-Home GeneralizationSeptember 17, 2026A person can walk into an unfamiliar home and start working immediately. Our understanding of the physical world carries from one environment to the next. We do not need to relearn how to make a bed because it is a different height, or how to tidy a room because the furniture is different. This is currently not true for robots. They learn the places they work in, one place at a time.Today we are introducing Helix 2.5, the most advanced neural network Figure has built. Helix 02 showed that a neural policy could coordinate a humanoid's whole body over long horizons, from unloading a dishwasher to running a logistics task autonomously for 200 hours.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at FigureAI.
Discussion
0 commentsMore from FigureAI
