Weekly must-read papers: GR00T 2, native VLA policies, agent guardrails and evals
TheTuringPost · x · 2026-11-11
The Turing Post curates this week's must-read papers, heavy on robotics and agents: GR00T 2, VFlash visual planning, La-JEPA compact robot policies, NEO-VLA native vision-language-action policies, RoboColosseum in-the-wild physical intelligence evals, and Open-HumanWAM. The attached image is Apple's HALO paper, which synthesizes vision-language-motion data across a thousand procedurally generated homes and transfers zero-shot to the real world with a 91% success rate over 169 trials. Agent-side picks include ToolSpace, LEANGUARD verified guardrails, Route Anything, agentic manipulation policies, plus Scientist World and FrontierPhysics eval environments. DreamX frames video diffusion as a universal latent action learner.
More from Embodied
- Waymo Pauses Freeway Robotaxi Rides in 4 Cities to Fix Construction Zone Handling — kscottz · 2026-11-11
- AI agents ran a biology lab end-to-end: AutoDNA completes DNA read-write cycle in 200 hours — CurieuxExplorer · 2026-11-11
- Cube-Solving Robot Demo Is Its Best Case: ~90 Moves per Solve on Average — haydendevs · 2026-11-11
- ETH researcher vibe-codes SuperMimic humanoid reproduction in 1h, LAFAN passes in 1.5 GPU-hours — ChongZzZhang · 2026-11-11
- Open-source companion robot Lamp ships at $699 with custom personality, skills and an app economy — dee_hw · 2026-11-11
- Must-read papers of the week: HALO humanoid motion, GR00T 2, VLA policies, agent tooling — TheTuringPost · 2026-11-11