Apple 提出 LEAD 方法，破解长程推理中的“不可恢复瓶颈”
LEAD: Breaking the No-Recovery Bottleneck in Long-Horizon Reasoning
Apple 研究发现，大语言模型在长程执行中即使有高层策略也不稳定，极端分解会导致“不可恢复瓶颈”——少数“困难”步骤上的持续错误变得不可逆转。为此提出 Lookahead-Enhanced Atomic Decomposition（LEAD），通过引入短程未来验证与聚合来打破这一瓶颈。该方法在受控算法谜题上验证了有效性。
Long-horizon execution in Large Language Models (LLMs) remains unstable even when high-level strategies are provided. Evaluating on controlled algorithmic puzzles, we demonstrate that while decomposition is essential for stability, extreme decomposition creates a “no-recovery bottleneck”. We show that this bottleneck becomes critical due to highly non-uniform error distribution, where consistent errors on a few “hard” steps become irreversible. To address this, we propose Lookahead-Enhanced Atomic Decomposition (LEAD). By incorporating short-horizon future validation and aggregating…
来源：Apple Machine Learning Research · machinelearning.apple.com