今日速览
| 序号 | 标题 | 来源 | 日期 | 主题 | 推荐等级 |
|---|---|---|---|---|---|
| 1 | ACEM: A Cost Estimation Model for Agentic Software Engineering | arXiv | 2026-08-03 | AI-Agent | 高 |
| 2 | RoMeRL: Balancing Feedback Coverage and the Memory-Reward Trap in Self-Evolving Agent Memory via Reduced-Order Utility States | arXiv | 2026-08-03 | RAG | 高 |
重点论文与技术动态
1. ACEM: A Cost Estimation Model for Agentic Software Engineering
- 来源：arXiv
- 日期：2026-08-03
- 作者/机构：Mohammad El-Ramly
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.SE
一句话结论
ACEM为代理软件工程开发了首个成本估算模型，整合了LLM token消耗、HITL监督和基础设施成本的非确定性特征。
核心内容
- ACEM将总代理开发成本分解为LLM、HITL和基础设施三个加性维度，解决传统模型依赖人类劳动的假设。
- 引入修订因子（RF）建模输出拒绝和重试的token冗余，上下文因子（CF）捕获上下文累积导致的token消耗上升，HITL强度评分（HIS）定义四级别监督分类。
- 映射用例点、故事点和功能点到token消耗估计，支持组织重用现有项目规划数据进行代理成本预测。
方法与数据
- 方法为模型结构和校准方法；数据：摘要未明确。
价值判断
- 值得关注：创新性处理代理软件工程中成本的非确定性现象，如相同任务token消耗波动和人类修正差异。
- 可复用点：直接利用现有软件工程指标（如故事点）估算代理成本，无需额外数据收集。
- 局限/待核查：模型常数需通过真实项目数据校准，当前为理论框架，依赖研究社区验证。
摘要
Traditional software cost estimation models, such as COCOMO II, Function Points, and Story Points, assume that development effort is primarily driven by human labor in design, coding, and testing. Agentic software engineering, where autonomous AI agents perform substantial implementation work and humans focus on planning, specification, and validation, challenges this assumption. New cost dimensions arise: large language model (LLM) token consumption across agent actions, Human-in-the-Loop (HITL) oversight effort, and infrastructure costs for agent orchestration and tooling. These costs are nondeterministic: identical tasks may consume different tokens, follow divergent reasoning paths, and require varying human correction, phenomena absent in traditional development. A new framework is needed to bridge standard sizing metrics with this cost structure. This paper proposes ACEM (Agentic Cost Estimation Model), which decomposes total agentic development cost into three additive dimensions: LLM, HITL, and infrastructure cost. ACEM introduces three constructs for agentic dynamics: the Revision Factor (RF), modeling token overhead from output rejection and retries; the Context Factor (CF), capturing rising token consumption as context accumulates; and the HITL Intensity Score (HIS), a four-level oversight classification scheme. It further maps Use Case Points, Story Points, and Function Points to estimated token consumption, enabling organizations to reuse existing project-scoping data for agentic cost forecasting. ACEM is presented as a fully specified model structure and calibration methodology, with constants left symbolic pending empirical grounding. As an early-stage proposal, it invites the research community to calibrate, test, and extend the model through real project data.2. RoMeRL: Balancing Feedback Coverage and the Memory-Reward Trap in Self-Evolving Agent Memory via Reduced-Order Utility States
- 来源：arXiv
- 日期：2026-08-03
- 作者/机构：Yi Yang, Zhennan Chen, Yihong Zhuang, Tiehan Fan, Yinan Chen, Jian Li
- 主题标签：
RAG,arXiv - 推荐等级：高
- 分类：cs.LG, cs.CL
一句话结论
RoMeRL通过减少阶效用状态有效平衡了反馈覆盖与记忆-奖励陷阱，显著提升自演化代理记忆的效率与性能。
核心内容
- 代理记忆系统面临两大挑战：轨迹索引效用随交互历史增长导致反馈分散在扩张状态空间，以及轨迹级奖励联合分配使无关经验进入记忆-奖励陷阱。
- RoMeRL采用固定维度的每任务记忆状态，通过结果极性和记忆动态分解效用空间，并利用固定语义坐标集动态更新整合新经验，集中反馈在有界效用支持。
- 实验在ALFWorld和LifelongAgentBench上验证，RoMeRL提升任务性能，降低Cold-Q比80.0%，增加反馈密度6.0倍，减少记忆维护大小84.4%，降低LLM调用21.1%。
方法与数据
- 方法：采用减少阶参数化，通过固定语义坐标集动态更新内容以集中反馈。
- 数据：基于ALFWorld和LifelongAgentBench基准测试评估。
价值判断
- 值得关注：解决自演化LLM代理记忆的核心挑战，实现高效反馈利用与持久奖励污染控制。
- 可复用点：减少阶效用状态框架可推广至其他自演化任务系统。
- 局限/待核查：摘要未明确提及局限。