xevals is now available↗
kamara
ManifestoResearchProjectsGitHubCommunity
  1. World-action models: planning in a learned latent

    Aug 20, 2026

    A VLA answers 'what should I do next?'. A world-action model answers 'what happens if I do this?' — and only the second question supports search, recovery, and calibrated refusal. We set out the latent dynamics objective, the rollout-horizon limit that governs it, and how a WAM composes with a VLA rather than replacing it.

    No. 02·8 min·WAM, World Models, Planning
  2. Vision-language-action models, from the inside out

    Jul 15, 2026

    A VLA is a vision-language model with the action space bolted onto its output head. We walk through the three design decisions that actually determine whether one works on a real arm — action tokenization, chunking, and the latency budget — and where each of them breaks.

    No. 01·8 min·VLA, Manipulation, Foundation Models
kamara

An open-source and open-science organization building infrastructure, evaluations, models, and research for Physical AI and its safety.

Organization

  • Manifesto

Work

  • Research
  • Projects
  • GitHub↗

Community

  • Discord↗
  • we@kamara.dev

Subscribe to Notes

New research issues, delivered when they publish. No noise.

© 2026 KamaraOpen source · Built in the open