CoRL 2026 to host 'Pretrain to Adapt' workshop: strongest robot policies aren't easiest to adapt
canondetortugas · x · 2026-09-09
CoRL 2026 will host a 'Pretrain to Adapt' workshop on November 12, 2026 in Austin, examining what makes a pretrained robot policy adaptable.
- The dominant paradigm pretrains a generalist manipulation policy on large diverse data, then adapts it via supervised/RL fine-tuning, in-context learning, or test-time steering
- Key tension: pretraining optimizes for broad competence, not for how readily a policy can be adapted — the strongest pretrained model isn't necessarily the best starting point, and there's no reliable way to tell before adaptation
- Existing adaptation methods typically treat the pretrained policy as fixed, ignoring how that choice limits downstream performance
- Submissions open via OpenReview, with a Slack community for discussion
More from Embodied
- RoboSim: open-source Rust robot-native simulation engine with physics, sensors and Python bindings — rsasaki0109 · 2026-09-09
- DriveZero: End-to-End Autonomous Driving Beyond Human Demonstrations — Hao He · 2026-09-09
- Enactic founder exits humanoid robotics startup after OpenArm open-source success — hiro_yams · 2026-09-09
- Apple acquires Sonera, betting on magnetic-field sensing to read brain activity in wearables — asymco · 2026-09-09
- You can now control a stepper motor for just $1.20 — _Stocko_ · 2026-09-09
- SimpleMemVLA feeds full video history to a VLM, beating dedicated memory modules for long-horizon robot manipulation — openbmb · 2026-09-09