用于时序推理的对比表示
Contrastive Representations for Temporal Reasoning
AI总结：
本文针对标准时序对比学习依赖虚假特征的问题，提出 CRTR 负采样表示方法，在 Sokoban 与魔方上实现强时序推理，并仅凭学习表示高效求解任意魔方状态。
摘要：
在经典 AI 中，感知依赖于学习基于状态的表示，而规划（可被视为对动作序列进行时序推理）通常通过搜索实现。我们研究这种推理是否能够反过来从同时捕捉感知结构和时序结构的表示中涌现。我们表明，标准的时序对比学习尽管广受欢迎，却往往因依赖虚假特征而无法捕捉时序结构。为解决这一问题，我们提出了 Combinatorial Representations for Temporal Reasoning（CRTR，用于时序推理的组合表示），该方法采用一种负采样方案，可证明地移除这些虚假特征并促进时序推理。CRTR 在具有复杂时序结构的领域取得了强劲结果，例如 Sokoban 和魔方。特别是在魔方任务中，CRTR 学到的表示能够在所有初始状态之间泛化，并使其能够以比 BestFS 更少的搜索步骤解开谜题，尽管所得解法更长。据我们所知，这是首个仅利用学习到的表示、而不依赖外部搜索算法，即可高效求解任意魔方状态的方法。
英文摘要：
In classical AI, perception relies on learning state-based representations, while planning, which can be thought of as temporal reasoning over action sequences, is typically achieved through search. We study whether such reasoning can instead emerge from representations that capture both perceptual and temporal structure. We show that standard temporal contrastive learning, despite its popularity, often fails to capture temporal structure due to its reliance on spurious features. To address this, we introduce Combinatorial Representations for Temporal Reasoning (CRTR), a method that uses a negative sampling scheme to provably remove these spurious features and facilitate temporal reasoning. CRTR achieves strong results on domains with complex temporal structure, such as Sokoban and Rubik's Cube. In particular, for the Rubik's Cube, CRTR learns representations that generalize across all initial states and allow it to solve the puzzle using fewer search steps than BestFS, though with longer solutions. To our knowledge, this is the first method that efficiently solves arbitrary Cube states using only learned representations, without relying on an external search algorithm.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
