今日速览
| 序号 | 标题 | 来源 | 日期 | 主题 | 推荐等级 |
|---|---|---|---|---|---|
| 1 | A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization | arXiv | 2026-10-08 | RAG | 高 |
| 2 | Use and Disuse: Intent-Structured Experience Consolidation for Memory and Learning in LLM Agents | arXiv | 2026-10-08 | AI-Agent | 高 |
重点论文与技术动态
1. A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization
- 来源：arXiv
- 日期：2026-10-08
- 作者/机构：Ming Chen, Rong-Xi Tan, Ke Xue, Yu-Jie Zhou, Taiye Lu, Zhi-Xuan Gao
- 主题标签：
RAG,arXiv - 推荐等级：高
- 分类：cs.LG, cs.AI, cs.NE
一句话结论
本文提出跨领域基准 AgenticBBO-Bench，系统评估了大模型智能体在黑盒优化（BBO）中的表现及关键设计因素的影响。
核心内容
- 智能体 BBO 在全部5个测试领域均优于直接 LLM 方法，并在4个领域超越了最佳数值优化器。
- 探究三大因素发现：增加数值工具未必持续提升性能；任务语义普遍有效但具体先验不太可靠；数值优化器能有效吸收智能体建立的搜索轨迹收益。
- 在前沿挑战中评测7种模型，发现 GPT-6 Astra 和 DeepSeek-V4.1-Flash 处于性能与成本的帕累托前沿。
方法与数据
- 采用统一有限预算评估协议与 Codex 框架，涵盖合成函数、超参数优化、数据库调优、芯片及分子设计5个领域的数据。
价值判断
- 值得关注：厘清了优化工具、先验知识及智能体搜索轨迹与数值优化器结合的实际效用。
- 可复用点：开源的 AgenticBBO-Bench 跨领域基准与评估代码。
- 局限/待核查：在1个领域未能超越最佳数值优化器，且摘要未明确具体失利领域及详细预算设置。
摘要
Black-box optimization (BBO) arises in many scientific and engineering problems where objective evaluations are expensive and limited. Recent large language model (LLM) agents offer a new way to approach BBO by combining task semantics, computation, optimization tools, and feedback-driven decision making, showing great potential due to the integration with mathematically rigorous tools. However, existing agentic BBO studies use different task domains and system configurations, making their results difficult to compare and the effects of individual design choices hard to isolate. We therefore introduce AgenticBBO-Bench, a cross-domain benchmark for agentic BBO spanning synthetic functions, hyperparameter optimization, database tuning, chip design, and molecular design under a unified finite-budget evaluation protocol. In our experiments, agentic BBO achieves higher family-averaged scores than direct LLM-based methods in all five domains and outperforms the best numerical optimizers in four. We further study three factors shaping agent performance: optimization tools, task information and prior knowledge, and the role of the LLM during search. Our results show that additional numerical tools do not consistently improve performance, task semantics are broadly useful while more specific priors are less reliable, and numerical optimizers can effectively absorb gains from search trajectories established by the agent. Finally, we introduce a five-task frontier challenge within AgenticBBO-Bench and evaluate seven LLMs under the Codex agent harness, where GPT-6 Astra and DeepSeek-V4.1-Flash lie on the Pareto frontier of performance and cost among the evaluated models. Our code is available at https://github.com/lamda-bbo/agentic-bbo.2. Use and Disuse: Intent-Structured Experience Consolidation for Memory and Learning in LLM Agents
- 来源：arXiv
- 日期：2026-10-08
- 作者/机构：Xiangyi Zeng, Baihang Liu, Xutong Wang, Ze Jin, Yunpeng Li, Qixu Liu
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.AI
一句话结论
本文提出分层记忆与持续学习架构Hippocam，通过基于意图的经验整合机制，使LLM智能体无需更新参数即可从自身经验中学习进化。
核心内容
- 嵌套意图：活跃上下文聚焦当前意图，已完成意图仅浓缩保留任务相关的结果与状态，舍弃完整工作细节。
- 递归整合：通过递归前缀整合使长期闲置经验逐渐抽象，同时保留原始交互以支持按需分层恢复细粒度细节。
- 动态更新：被召回并重新融入当前工作的历史经验，会与新经验共同经历后续整合，实现强化、补充和更新。
方法与数据
- 方法：提出受人类记忆启发的Hippocam架构，通过“用进废退”动态机制打通工作上下文、长期记忆与技能学习。
- 数据：摘要未明确。
价值判断
- 值得关注：无需更新参数即可将连续经验转化为可复用知识，支持智能体长期自主运行。
- 可复用点：嵌套意图上下文压缩，及“闲置抽象、按需恢复”的记忆整合机制。
- 局限/待核查：摘要未明确具体的实验数据集与性能指标。