paper-with-me

홈 › Papers

Marginal Advantage Accumulation for Memory-Driven Agent Self-Evolution

2026-06-18 · Mingyu Yang, Keye Zheng, Congchao Cheng, Yujie Liu, Xingkang Lu, Fan Jiang, Yefei Zheng arxiv

In batch-style trace distillation, the same memory operation may receive contradictory feedback across different batches. Existing methods lack a cross-batch, operation-level evidence accumulation mechanism, making it impossible to distinguish stably effective operations from accidental hits. This paper formalizes the requirement as two structural conditions, alignability and comparability, and proposes Marginal Advantage Accumulation (MAA). MAA constructs differential signals to make them comparable across batches, accumulates signed evidence per operation via EMA, and ensures cross-batch traceability through semantic identity merging. As a post-processing architecture, MAA achieves the best results in 14 out of 16 settings across 4 benchmarks and 4 target models, consistently outperforming existing batch-level distillation baselines and matching or surpassing online alternatives in most settings, while reducing optimization-phase token consumption by approximately 75%.

📄 PDF Abstract BibTeX arXiv:2606.20475

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

AI Agents Need Memory Control Over More Context

2026-01-15 · Fouad Bousetouane arxiv

AI agents are increasingly used in long, multi-turn workflows in both research and enterprise settings. As interactions grow, agent behavior often degrades due to loss of constraint focus, error accumulation, and memory-…

Remember Me, Refine Me: A Dynamic Procedural Memory Framework for Experience-Driven Agent Evolution

2025-12-11 · Zouying Cao, Jiaji Deng, Li Yu, Weikang Zhou 외 arxiv

Procedural memory enables large language model (LLM) agents to internalize "how-to" knowledge, theoretically reducing redundant trial-and-error. However, existing frameworks predominantly suffer from a "passive accumulat…

Empirical-MCTS: Continuous Agent Evolution via Dual-Experience Monte Carlo Tree Search

2026-02-04 · Hao Lu, Haoyuan Huang, Yulin Zhou, Chen Li 외 arxiv

Inference-time scaling strategies, particularly Monte Carlo Tree Search (MCTS), have significantly enhanced the reasoning capabilities of Large Language Models (LLMs). However, current approaches remain predominantly sta…

Darwinian Memory: A Training-Free Self-Regulating Memory System for GUI Agent Evolution

2026-01-30 · Hongze Mi, Yibo Feng, WenJie Lu, Song Cao 외 arxiv

Multimodal Large Language Model (MLLM) agents facilitate Graphical User Interface (GUI) automation but struggle with long-horizon, cross-application tasks due to limited context windows. While memory systems provide a vi…

MetaEvo: A Meta-Optimization Framework for Experience-Driven Agent Evolution

2026-05-29 · Bowen Ren, Heyan Huang, Yinghao Li, Yang Gao arxiv

Large language models (LLMs) exhibit strong reasoning capabilities, yet most LLM-based agents are statically deployed and unable to improve through task interactions. Existing experience-driven methods often rely on memo…