跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.22491 · v1

Progress Predicts Policy

首次提交 2026-11-17 17:46
版本更新 2026-11-17 17:46
雷达观测 2026-11-18 11:33
arXiv ↗PDF ↗

摘要

A task specifies how the world should change, while a policy specifies what the robot should do. We introduce Progress Predicts Policy (PPP), which learns a language-conditioned scalar field through the commands induced by its differential. We obtain action conditioning by composing a latent progress field with a shared world model and read its action gradient at a fixed reference command. Continuous commands scale this gradient by a learned positive gain, while discrete commands use finite differences of the same scalar. This yields a hybrid policy without a separate action head or iterative search. Fitting demonstration progress values alone does not ensure that action gradients align with expert commands. PPP trains the induced commands by imitation and uses double backpropagation for the continuous gradient readout to shape the field's local control geometry. Demonstration timing calibrates field values on observed states without additional annotation. The continuous readout remains learnable without progress labels but these timing-derived targets support progress monitoring during execution. Experiments in simulated and real robot manipulation show that PPP outperforms behavioral cloning under a shared architecture.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-17 17:46 · 最新观测版本

Progress Predicts Policy

A task specifies how the world should change, while a policy specifies what the robot should do. We introduce Progress Predicts Policy (PPP), which learns a language-conditioned scalar field through the commands induced by its differential. We obtain action conditioning by composing a latent progress field with a shared world model and read its action gradient at a fixed reference command. Continuous commands scale this gradient by a learned positive gain, while discrete commands use finite differences of the same scalar. This yields a hybrid policy without a separate action head or iterative search. Fitting demonstration progress values alone does not ensure that action gradients align with expert commands. PPP trains the induced commands by imitation and uses double backpropagation for the continuous gradient readout to shape the field's local control geometry. Demonstration timing calibrates field values on observed states without additional annotation. The continuous readout remains learnable without progress labels but these timing-derived targets support progress monitoring during execution. Experiments in simulated and real robot manipulation show that PPP outperforms behavioral cloning under a shared architecture.

查看此版本解读 →查看此版本原文 ↗