让大模型在测试时自适应，解决知识过时与组合失效问题
PRECEPT: Planning Resilience via Experience, Context Engineering & Probing Trajectories A Unified Framework for Test-Time Adaptation with Compositional Rule Learning and Pareto-Guided Prompt Evolution
- 通过结构化条件精确匹配规则，避免语义误解
- 融合贝叶斯可靠性判断与阈值失效机制，应对知识冲突与漂移
- 基于帕累托优化的提示进化，显著提升推理效率与鲁棒性
大型语言模型代理将知识以自然语言存储，在条件数量增加时面临严重检索退化，难以可靠组合已学规则，且缺乏检测过时或对抗性知识的显式机制。本文提出PRECEPT，一个统一的测试时自适应框架，包含三个紧密耦合组件：（1）基于结构化条件键的确定性精确匹配规则检索；（2）考虑冲突感知的记忆机制，结合贝叶斯源可靠性与阈值驱动的规则失效机制；（3）COMPASS，一种基于帕累托引导的提示演化外循环。精确检索消除了确定路径上的部分匹配解释错误（构造上为0%，而独立模型下N=10时达94.4%），支持通过语义层级堆叠进行组合；冲突感知记忆可解决静态-动态不一致并支持漂移适应；COMPASS通过相同端到端执行流程评估提示。实验结果（9–10个种子）显示：相较于Full Reflexion，PRECEPT首次尝试优势提升41.1个百分点（d>1.9），组合泛化能力提升33.3个百分点（d=1.55），两路物流组合任务中P₁达到100%（d=2.64），持续学习增益40–55个百分点，对抗性静态知识下仍保持100%物流性能（整合任务部分恢复），漂移恢复能力提升55.0个百分点（d=0.95，p=0.031），步骤数减少61%。核心对比均具统计显著性，多数p<0.001。
原文摘要 · Abstract (English)
LLM agents that store knowledge as natural language suffer steep retrieval degradation as condition count grows, often struggle to compose learned rules reliably, and typically lack explicit mechanisms to detect stale or adversarial knowledge. We introduce PRECEPT, a unified framework for test-time adaptation with three tightly coupled components: (1) deterministic exact-match rule retrieval over structured condition keys, (2) conflict-aware memory with Bayesian source reliability and threshold-based rule invalidation, and (3) COMPASS, a Pareto-guided prompt-evolution outer loop. Exact retrieval eliminates partial-match interpretation errors on the deterministic path (0% by construction, vs 94.4% under Theorem~B.6's independence model at N=10) and supports compositional stacking through a semantic tier hierarchy; conflict-aware memory resolves static--dynamic disagreements and supports drift adaptation; COMPASS evaluates prompts through the same end-to-end execution pipeline. Results (9--10 seeds): PRECEPT achieves a +41.1pp first-try advantage over Full Reflexion (d>1.9), +33.3pp compositional generalization (d=1.55), 100% $P_1$ on 2-way logistics compositions (d=2.64), +40--55pp continuous learning gains, strong eventual robustness under adversarial static knowledge (100% logistics with adversarial SK active; partial recovery on integration), +55.0pp drift recovery (d=0.95, p=0.031), and 61% fewer steps. Core comparisons are statistically significant, often at p<0.001.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。