← 研踪 PaperTrailarXiv:2605.16725 · v2
Baba in Wonderland: Online Self-Supervised Dynamics Discovery for Executable World Models
首次提交 2026-05-16 08:18
版本更新 2026-11-04 19:01
雷达观测 2026-11-05 11:10
摘要
Executable world models can be read, edited, executed, and reused for planning, but only if the program captures the environment's transition law rather than semantic shortcuts in its surface vocabulary. We study online executable world-model learning under prior misalignment, where an agent must induce state-dependent dynamics from interaction evidence alone, without rule descriptions, reward signals, or trustworthy lexical priors. We introduce Alice, a closed-loop system that treats failed candidate updates as structural signal: when a candidate explains a new transition but loses previously explained ones, the preservation conflict reveals dynamics that the current program had conflated. Alice refines these conflicts into hypothesis classes that both provide a few broken examples, one per affected class, for update and guide frontier exploration toward transitions that are novel and underrepresented with respect to the current program. We evaluate Alice on Baba in Wonderland, a prior-misaligned variant of Baba Is You that preserves simulator dynamics while replacing semantically meaningful rule-property labels with unrelated words. Experiments show that Alice substantially improves executable world-model learning under prior misalignment, and five-run ablations support overall improvement trends as class structure and exploration signals are added. Project page: https://ssw03270.github.io/BABAinWonderland_NeurIPS-26_Page/.
作者备注：NeurIPS 2026 camera-ready version; revised manuscript with final corrections
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。
v2 · 2026-11-04 19:01 · 最新观测版本
Baba in Wonderland: Online Self-Supervised Dynamics Discovery for Executable World Models
Executable world models can be read, edited, executed, and reused for planning, but only if the program captures the environment's transition law rather than semantic shortcuts in its surface vocabulary. We study online executable world-model learning under prior misalignment, where an agent must induce state-dependent dynamics from interaction evidence alone, without rule descriptions, reward signals, or trustworthy lexical priors. We introduce Alice, a closed-loop system that treats failed candidate updates as structural signal: when a candidate explains a new transition but loses previously explained ones, the preservation conflict reveals dynamics that the current program had conflated. Alice refines these conflicts into hypothesis classes that both provide a few broken examples, one per affected class, for update and guide frontier exploration toward transitions that are novel and underrepresented with respect to the current program. We evaluate Alice on Baba in Wonderland, a prior-misaligned variant of Baba Is You that preserves simulator dynamics while replacing semantically meaningful rule-property labels with unrelated words. Experiments show that Alice substantially improves executable world-model learning under prior misalignment, and five-run ablations support overall improvement trends as class structure and exploration signals are added. Project page: https://ssw03270.github.io/BABAinWonderland_NeurIPS-26_Page/.
查看此版本解读 →查看此版本原文 ↗