MIT's FLEX pipeline lets robots refine task representations from natural language corrections
andreea7b · x · 2026-11-11
At CoRL 2026, MIT CSAIL researchers (Audrey Lee, Nathaniel Dennler, Ilia Sucholutsky, Andreea Bobu) present FLEX, a Flexible Feature Editing Pipeline that uses LLMs to iteratively refine a robot's internal task representation from natural language corrections via three operations: add missing features, generalize overly specific ones, and mask irrelevant ones.
Trajectory demos show robots how to act but not why, often producing misaligned representations. Evaluated on 2D navigation, 7-DoF Franka tabletop manipulation, and a real-robot study with naturalistic human corrections, FLEX consistently beats single-operation and unscaffolded baselines in normalized reward while keeping representations compact and transferable. Poster: Nov 10, CoRL Session 2 #37.
More from Embodied
- Supersplat XR update: Quest 3 streamed scenes jump from 9.7 to 30+ fps, Vision Pro gets WebGPU — willeastcott · 2026-11-11
- Cartwheel's thesis: generative kinematics atop a 50Hz tracking policy, no teleop grind — andrew_n_carr · 2026-11-11
- Cartwheel releases three open motion-tracking policies for humanoid robots on Hugging Face — andrew_n_carr · 2026-11-11
- Claude Fable 5.2 designs a humanoid robot overnight with CAD, sim and reel — davidpattersonx · 2026-11-11
- Russ Tedrake's CoRL talk: deploy robots early, train on intervention data to scale — DominiqueCAPaul · 2026-11-11
- La-JEPA: Huawei-led 1B robot policy beats 12.4B Fast-WAM (60.8% vs 39.9%) and runs 5x faster — burny_tech · 2026-11-11