Three seconds to learn: Gen1.5's few-shot skill learning may hinge on one attention masking rule
KyeGomezB · x · 2026-08-20
GEN is the first model line where one-shot and few-shot learning of physical skills appears to emerge from pretraining alone — three seconds of demonstration and it performs the task with no gradient updates. The author hypothesizes the key is a simple masking rule: an action token at time t may only attend to sensor tokens at t−δ, where δ is the inference latency sampled during training. This could explain how GEN avoids a System 1/System 2 split and inference-time guidance while hitting 100 Hz with a 7B model.
More from Embodied
- MIT's C-GMT gives humanoid robots whole-body compliance without sacrificing motion tracking — MIT_CSAIL · 2026-10-16
- OpenRoboto, Bittensor's largest robotics subnet, puts its data collectors on Solana via peaqOS — bittingthembits · 2026-10-16
- Robotics data tokenization landscape: Khala maps peaq, THOR, BitRobot and Axis — 0xSammy · 2026-10-16
- Perseverance rover logs 28.1 miles on Mars, breaking Opportunity's record in half the time — BrettKrieger12 · 2026-10-16
- Trossen bimanual arms portion food in commercial kitchen with models fine-tuned via Physical Intelligence API — kscottz · 2026-10-16
- Safeworld emerges from stealth with $12M to build 'CI/CD for robots' safety testing — lukas_m_ziegler · 2026-10-16