Team uses Claude Opus 5.5 to reproduce core claims of 10 recent robotics papers end to end

teortaxesTex · x · 2026-10-16

• Nicolas Keller's team handed Claude Opus 5.5 their datasets, training infra, and robot farm, asking it to reproduce core claims of 10 recent papers, some only weeks old.
• The agent runs the full workflow end to end — data curation, annotation, hardware setup, real-world evaluation — while humans review and steer.
• Motivation: reproducibility in robotics is extremely costly even with open code, and verifying papers at scale is infeasible for humans.
• The effort hinges on rigorous benchmarking; details shared in a 17-post thread.

Original post →

More from Embodied

• EmbodiedEvo tops RoboDojo as agents inherit reusable skills and lessons — Aiden_Tech_Ai · 2026-10-16
• Sunday Robotics is hiring a Senior Infra PM to scale consumer robot deployments — saranormous · 2026-10-16
• Bond's Submarine Lotus Esprit Spotted at Event, Fueling Steering-Wheel-Free Robotaxi Speculation — PTrubey · 2026-10-16
• Maker wires Grok and Claude into ROS2 robot with RealSense stereo cameras — chrismatthieu · 2026-10-16
• NEURA Robotics launches Neuron AI, a unified Perception-Language-Action model for robots — Distinct-Question-16 · 2026-10-16
• Dev demos open wearable AI passport running multilingual flashcards on tiny memory — ezshine · 2026-10-16

Embodied channel →