REVIEW 3 major objections 5 minor 11 cited by
G-Memory: Tracing Hierarchical Memory for Multi-Agent Systems
T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read A three-tier graph memory lets LLM-based multi-agent teams learn from past collaborations, lifting success by up to 20.89% on embodied tasks and 10.12% on knowledge QA without touching the base framework.
desk verdict A well-motivated hierarchical memory architecture for MAS with a broad evaluation, but the empirical claims are undermined by missing error bars and test-set hyperparameter selection. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is the three-tier graph hierarchy with bi-directional traversal. Coarse similarity search over the query graph returns candidate past tasks; a one-hop expansion widens the candidate set; an upward traversal from query to insight graph collects high-level lessons, and a downward traversal uses an LLM-facilitated sparsifier to condense each relevant interaction graph into the dialogue segments that matter for the current task. An operator then personalizes these memories for each agent role, and after execution the hierarchy is updated: new interaction graphs are stored, new query nodes are linked to the relevant past queries and to the queries supporting the used insights, and new or merged insights are distilled from contrasting successful and failed trajectories. The hierarchy is what lets the system trade off abstraction and detail: general lessons guide strategy, while condensed trajectories provide concrete corrective steps.
What would settle it
Run the identical comparison with a held-out test set: build all G-Memory graphs using only queries from a training split, then evaluate on a disjoint split (or on a second distribution entirely). If the improvements over the no-memory baseline shrink to near zero or reverse, the central claim of cross-trial generalization fails. A cheaper check is to clear memory between trials and confirm success rates fall back to baseline.
Extended reading notes
Core claim
The central claim is that the bottleneck in multi-agent self-evolution is memory, not orchestration. G-Memory counters the simplistic designs it surveys with a three-tier graph: the insight graph holds distilled, generalizable lessons together with the queries that support them; the query graph stores past queries, their success or failure status, and a semantic network connecting them; the interaction graph records every agent utterance and the influence edges that trace how one utterance inspired another. For a new query, G-Memory retrieves similar past queries, expands to their one-hop neighbors, then traverses upward to pull relevant insights and downward to condense the most useful interaction histories into core subgraphs. Each agent receives memory filtered to its role before the team acts, and after execution the query and insight graphs are extended with the new experience. The paper reports that this cycle consistently improves performance across three task domains, with the largest gains on embodied action tasks.
Load-bearing premise
The reported gains assume that the memory accumulated during earlier evaluation trials does not leak information about later test queries; the experimental setup does not state a train/test split, so the improvements could reflect within-distribution memorization rather than cross-trial generalization.
Editorial extensions
If this is right
- G-Memory can be retrofitted onto any MAS that accepts extra context: the reported gains are obtained without modifying the frameworks themselves.
- Memory should be role-specific: giving every agent the same retrieved history underperforms, suggesting that agents with different functions need different cues.
- High-level insights and fine-grained interaction traces are complementary; removing either one costs roughly 3 to 5 percentage points on the tested frameworks.
- Teams improve with exposure: on embodied benchmarks, G-Memory reaches higher success ceilings and succeeds with fewer trials as memory accumulates.
- The performance gain is not bought by excessive prompting: token consumption stays in line with cheaper baselines.
Reading between the lines
- A decisive untested question is generalization beyond the evaluation distribution: the paper does not report a train/test split, and its success curves show memory accumulating during the test run; a held-out evaluation would separate cross-trial learning from memorization.
- The insight graph's cross-task edges hint at a transfer extension the paper does not run: checking whether lessons learned in one domain (such as verifying web-search facts) improve performance in another (such as embodied object placement).
- The query graph could support richer retrieval keys than semantic similarity, such as failure type or agent role; the paper only explores embedding similarity plus one-hop expansion.
- Because insight distillation and trajectory sparsification are done by LLMs, the quality of the memory likely depends on the backbone's summarization reliability; the paper does not isolate this effect.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript introduces G-Memory, a hierarchical three-tier graph memory system for LLM-based multi-agent systems (MAS), comprising an insight graph, a query graph, and an interaction graph. The method performs coarse-grained retrieval over the query graph, bi-directional traversal to retrieve high-level insights and condensed interaction trajectories, and then updates all three levels after task execution. The authors claim that G-Memory is plug-and-play and improves embodied action success by up to 20.89% and knowledge QA accuracy by up to 10.12% across five benchmarks, three LLM backbones, and three MAS frameworks. The evaluation is currently reported as single point estimates without error bars, with hyperparameters selected by inspecting the same benchmarks whose results are reported, and without an explicit train/test split in the memory-accumulation protocol.
Significance. If the empirical claims withstand scrutiny, G-Memory addresses a real and currently underserved need: most existing MAS memory mechanisms are limited to inside-trial memory or coarse final artifacts, whereas G-Memory offers role-specific, multi-granularity memory with a principled graph-based organization. The architecture is clear and coherent, the ablations in Figure 4c show that both the insight and interaction components contribute, and the paper covers a wide range of benchmarks, backbones, and frameworks. The public code availability is a further strength. However, the central contribution is empirical, and the current evidence is not yet sufficient to support the headline numbers: there are no variance estimates, hyperparameters are tuned on the same test benchmarks, and the evaluation does not clarify whether memory is built from queries disjoint from those used for final evaluation. The use of LLMs to generate insights is not circular because performance is measured on external benchmarks, but the statistical and protocol issues are load-bearing for the paper's main claim.
major comments (3)
- [§5.1, §5.4, Tables 1–3] The hyperparameters k and hop are selected by observing performance on the same five benchmarks whose results are then reported as the headline numbers. Section 5.4 states that 'we employ 1-hop expansion and k∈{1,2} throughout the experiments' after Figure 4 shows performance on ALFWorld, PDDL, and FEVER for different hop counts and k values. This is a form of test-set tuning: the reported 20.89% and 10.12% gains reflect configurations chosen on the benchmark test sets. Please provide an independent validation split for hyperparameter selection, report results for all configurations, or otherwise justify that the choice does not inflate the reported gains.
- [§5.1, §A, Tables 1–3] The experiments are reported as single point estimates with no number of runs, random seeds, sampling temperatures, or confidence intervals. Both LLM decoding and MacNet's random graph topology are stochastic, so differences of several percentage points between rows in Tables 1–3 may be consistent with sampling noise. For example, Table 3 shows Voyager on DyLAN/HotpotQA at 30.71 versus no-memory at 30.61, a 0.10-point difference, while other comparisons range from a few points to more than twenty points. Without multiple seeds and variance estimates, the reader cannot determine which improvements are meaningful. Please report means and standard errors over repeated runs, or an equivalent significance analysis.
- [§5, §B.1, Figures 6a–c] The evaluation protocol never specifies a train/test split. Figures 6a–c show success curves over trials in which memory is accumulated during the evaluation run, and Section A does not state whether the queries used to build memory are disjoint from the queries whose success is reported. If memory is built from the same distribution being tested, the observed gains could reflect within-distribution memorization rather than cross-trial generalization, which is what the abstract's 'generalizable insights' claims. Please clarify the split between memory-building queries and evaluation queries, or add a held-out evaluation where memory is constructed only on a training set.
minor comments (5)
- [Appendix C] The heading 'Inisght Summarization Function' contains a typo; it should be 'Insight Summarization Function'.
- [Eq. (9)] The set QR is used in Eq. (9) but is not explicitly named when the top-M relevant queries are introduced in Eq. (7); please define QR there or restate it in the text around Eq. (9).
- [§5.4] The parameter M in Eq. (7) is listed as being chosen from {2,3,4,5} in Section 5.1, but Section 5.4 reports sensitivity analysis only for hop and k, not for M. Please state how M was set and whether the results are robust to that choice.
- [§5.4] The text says 'the optimal k is among {1,2}' and later 'we employ 1-hop expansion and k∈{1,2} throughout the experiments,' which is ambiguous about whether k=1 or k=2 is selected per benchmark, per framework, or fixed globally. Please state the exact setting for each reported cell.
- [Abstract and Introduction] The phrase 'we are alarmed to discover' is editorializing and out of place in a formal research paper; please rephrase.
Circularity Check
No circularity found: the reported gains are empirical comparisons on external benchmarks, and the memory mechanism is not defined in terms of the measured outcomes.
full rationale
The paper's central claim is an empirical performance comparison against five external benchmarks (ALFWorld, SciWorld, PDDL, HotpotQA, FEVER) under fixed MAS frameworks and LLM backbones. G-Memory's equations (4)-(11) define a retrieval and update mechanism; they do not encode the evaluation metrics or presuppose the reported success rates. The insight, query, and interaction graphs are constructed from trajectories and updated with environmental feedback, and the headline improvements are measured by external task success/accuracy, not by a quantity defined in terms of G-Memory's own outputs. No parameter is fitted to a target label and then reported as a prediction: the choices of k and hop are selected in Section 5.4 on the same benchmarks, which is a test-set tuning concern, not a reduction of the result to an input. Self-citations ([30], [51], [69], [71]) appear only in related-work discussion and are not used to justify the architecture or the empirical claim. The fact that LLM-generated insights are later consumed by LLM agents is an architectural choice, not a circular definition. Therefore no circular step is exhibited and the score is 0.
Assumptions & free parameters
free parameters (3)
- top-k retrieved queries (k) =
1 or 2
- number of sparsified interaction graphs (M) =
2 to 5
- hop expansion (hop) =
1
assumptions (3)
- domain assumption Historical trajectories of similar queries contain transferable guidance for new queries.
- domain assumption LLM-based summarization and sparsification preserve critical information without distortion.
- domain assumption The evaluation benchmarks' test tasks are independent of the memory-building trials.
Cite this review
Pith. "Pith review of G-Memory: Tracing Hierarchical Memory for Multi-Agent Systems." pith.science (2026). https://pith.science/paper/WZRYOMZY
@misc{pith2026250607398,
author = {Pith},
title = {Pith review of: G-Memory: Tracing Hierarchical Memory for Multi-Agent Systems},
year = {2026},
howpublished = {\url{https://pith.science/paper/WZRYOMZY}},
note = {Machine review of arXiv:2506.07398}
}
abstract
Large language model (LLM)-powered multi-agent systems (MAS) have demonstrated cognitive and execution capabilities that far exceed those of single LLM agents, yet their capacity for self-evolution remains hampered by underdeveloped memory architectures. Upon close inspection, we are alarmed to discover that prevailing MAS memory mechanisms (1) are overly simplistic, completely disregarding the nuanced inter-agent collaboration trajectories, and (2) lack cross-trial and agent-specific customization, in stark contrast to the expressive memory developed for single agents. To bridge this gap, we introduce G-Memory, a hierarchical, agentic memory system for MAS inspired by organizational memory theory, which manages the lengthy MAS interaction via a three-tier graph hierarchy: insight, query, and interaction graphs. Upon receiving a new user query, G-Memory performs bi-directional memory traversal to retrieve both $\textit{high-level, generalizable insights}$ that enable the system to leverage cross-trial knowledge, and $\textit{fine-grained, condensed interaction trajectories}$ that compactly encode prior collaboration experiences. Upon task execution, the entire hierarchy evolves by assimilating new collaborative trajectories, nurturing the progressive evolution of agent teams. Extensive experiments across five benchmarks, three LLM backbones, and three popular MAS frameworks demonstrate that G-Memory improves success rates in embodied action and accuracy in knowledge QA by up to $20.89\%$ and $10.12\%$, respectively, without any modifications to the original frameworks. Our codes are available at https://github.com/bingreeky/GMemory.
Figures
Figures from the paper (7 more)
Forward citations
Cited by 11 Pith papers
-
Knowledge-Centric Self-Improvement
A protocol that keeps agents generic and disposable while improving a persistent curated knowledge base outperforms agent-centric self-improvement baselines on ARC, coding, and terminal benchmarks.
-
Memory as a Controlled Process: Learned Adaptive Memory Management for LLM Agents
A tabular UCB controller trained on task success improves LLM-agent memory use over fixed heuristics, without extra LLM calls.
-
Mosaic: Runtime-Efficient Multi-Agent Embodied Planning
Mosaic delivers 27-32% faster multi-agent embodied execution and 4-10 point higher success via agent-centric relative memory plus per-step ILP action allocation.
-
What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents
KV-cache eviction, prompt compression, recurrent state bounding, and agent memory consolidation are unified as one rate-distortion problem with a shared lower bound, shared failure mode, and transferable mechanisms.
-
AgentKVShift: Efficient KV Cache Reuse for Agentic Memory Systems
AgentKVShift corrects all reused KV cache entries in agentic memory by estimating a chunk-level residual offset from a small probe set, recovering near-fresh accuracy at 10–30% recompute ratios.
-
ActMem: Bridging the Gap Between Memory Retrieval and Reasoning in LLM Agents
ActMem stores agent dialogue as a causal and semantic knowledge graph and uses LLM counterfactual reasoning to retrieve implicit constraints, outperforming RAG-style baselines on a new reasoning benchmark.
-
Cost and Accuracy of Long-Term Memory in Distributed Multi-Agent Systems Based on Large Language Models
A two-framework testbed comparison claims mem0 is Pareto-optimal over Graphiti for distributed LLM agents because its lower cost is paired with accuracy that is not significantly different.
-
RGMem: Renormalization Group-inspired Memory Evolution for Language Agents
RGMem, a hierarchical memory framework with thresholded updates inspired by renormalization group ideas, reports state-of-the-art scores on the LOCOMO long-term conversational memory benchmark.
-
Meta-Policy Reflexion: Reusable Reflective Memory and Rule Admissibility for Resource-Efficient LLM Agent
Storing LLM reflections as predicate-like rules and applying them via prompt guidance plus hard validity filtering is reported to improve AlfWorld task accuracy over Reflexion, without weight updates.
-
Multi-level Value Alignment in Agentic AI Systems: Survey and Perspectives
A survey proposes a macro-meso-micro value framework for agentic AI alignment and maps applications, methods, and benchmarks onto it.
-
Token-Operations-Oriented Inference Optimization Techniques for Large Models
The paper introduces a four-layer technical architecture for token-operations-oriented inference optimization in large models and reviews key technologies and industry status at each layer.
Reference graph
Works this paper leans on
-
[1]
Organizational memory.Academy of management review, 16(1):57–91, 1991
James P Walsh and Gerardo Rivera Ungson. Organizational memory.Academy of management review, 16(1):57–91, 1991
1991
-
[2]
Palm-e: An embodied multimodal language model
Danny Driess, Fei Xia, Mehdi SM Sajjadi, Corey Lynch, Aakanksha Chowdhery, Ayzaan Wahid, Jonathan Tompson, Quan Vuong, Tianhe Yu, Wenlong Huang, et al. Palm-e: An embodied multimodal language model. 2023
2023
-
[3]
Shihao Wang, Zhiding Yu, Xiaohui Jiang, Shiyi Lan, Min Shi, Nadine Chang, Jan Kautz, Ying Li, and Jose M Alvarez. Omnidrive: A holistic llm-agent framework for autonomous driving with 3d perception, reasoning and planning. arXiv preprint arXiv:2405.01533, 2024
arXiv 2024
-
[4]
Steve-eye: Equipping llm-based embodied agents with visual perception in open worlds
Sipeng Zheng, Jiazheng Liu, Yicheng Feng, and Zongqing Lu. Steve-eye: Equipping llm-based embodied agents with visual perception in open worlds. arXiv preprint arXiv:2310.13255, 2023
arXiv 2023
-
[5]
Editable scene simulation for autonomous driving via collaborative llm-agents
Yuxi Wei, Zi Wang, Yifan Lu, Chenxin Xu, Changxing Liu, Hao Zhao, Siheng Chen, and Yanfeng Wang. Editable scene simulation for autonomous driving via collaborative llm-agents. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15077–15087, 2024
2024
-
[6]
Knowagent: Knowledge-augmented planning for llm-based agents
Yuqi Zhu, Shuofei Qiao, Yixin Ou, Shumin Deng, Shiwei Lyu, Yue Shen, Lei Liang, Jinjie Gu, Huajun Chen, and Ningyu Zhang. Knowagent: Knowledge-augmented planning for llm-based agents. arXiv preprint arXiv:2403.03101, 2024
arXiv 2024
-
[7]
Plan-and-act: Improving planning of agents for long-horizon tasks
Lutfi Eren Erdogan, Nicholas Lee, Sehoon Kim, Suhong Moon, Hiroki Furuta, Gopala Anu- manchipalli, Kurt Keutzer, and Amir Gholami. Plan-and-act: Improving planning of agents for long-horizon tasks. arXiv preprint arXiv:2503.09572, 2025
arXiv 2025
-
[8]
Understanding the planning of llm agents: A survey
Xu Huang, Weiwen Liu, Xiaolong Chen, Xingmei Wang, Hao Wang, Defu Lian, Yasheng Wang, Ruiming Tang, and Enhong Chen. Understanding the planning of llm agents: A survey. arXiv preprint arXiv:2402.02716, 2024
arXiv 2024
Show all 85 references
-
[9]
Agent q: Advanced reasoning and learning for autonomous ai agents
Pranav Putta, Edmund Mills, Naman Garg, Sumeet Motwani, Chelsea Finn, Divyansh Garg, and Rafael Rafailov. Agent q: Advanced reasoning and learning for autonomous ai agents. arXiv preprint arXiv:2408.07199, 2024
2024 arXiv
-
[10]
The landscape of emerging ai agent architectures for reasoning, planning, and tool calling: A survey
Tula Masterman, Sandi Besen, Mason Sawtell, and Alex Chao. The landscape of emerging ai agent architectures for reasoning, planning, and tool calling: A survey. arXiv preprint arXiv:2404.11584, 2024
2024 arXiv
-
[11]
Embodied agent interface: Benchmarking llms for embodied decision making
Manling Li, Shiyu Zhao, Qineng Wang, Kangrui Wang, Yu Zhou, Sanjana Srivastava, Cem Gokmen, Tony Lee, Erran Li Li, Ruohan Zhang, et al. Embodied agent interface: Benchmarking llms for embodied decision making. Advances in Neural Information Processing Systems, 37:100428–100534, 2024
2024
-
[12]
Embodied multi-modal agent trained by an llm from a parallel textworld
Yijun Yang, Tianyi Zhou, Kanxue Li, Dapeng Tao, Lusong Li, Li Shen, Xiaodong He, Jing Jiang, and Yuhui Shi. Embodied multi-modal agent trained by an llm from a parallel textworld. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 26275...
2024
-
[13]
Autogen: Enabling next-gen llm applications via multi-agent conversation framework, August 01, 2023 2023
Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Shaokun Zhang, Erkang Zhu, Beibin Li, Li Jiang, Xiaoyun Zhang, and Chi Wang. Autogen: Enabling next-gen llm applications via multi-agent conversation framework, August 01, 2023 2023. 10
2023
-
[14]
Daya Guo, Qihao Zhu, Dejian Yang, Zhenda Xie, Kai Dong, Wentao Zhang, Guanting Chen, Xiao Bi, Y . Wu, Y . K. Li, Fuli Luo, Yingfei Xiong, and Wenfeng Liang. Deepseek-coder: When the large language model meets programming – the rise of code intelligence, 2024
2024
-
[15]
Data interpreter: An llm agent for data science
Sirui Hong, Yizhang Lin, Bang Liu, Bangbang Liu, Binhao Wu, Ceyao Zhang, Chenxing Wei, Danyang Li, Jiaqi Chen, Jiayi Zhang, et al. Data interpreter: An llm agent for data science. arXiv preprint arXiv:2402.18679, 2024
2024 arXiv
-
[16]
V oyager: An Open-Ended Embodied Agent with Large Language Models
Guanzhi Wang, Yuqi Xie, Yunfan Jiang, Ajay Mandlekar, Chaowei Xiao, Yuke Zhu, Linxi Fan, and Anima Anandkumar. V oyager: An Open-Ended Embodied Agent with Large Language Models. arXiv e-prints, page arXiv:2305.16291, May 2023
2023 arXiv
-
[17]
Driving with llms: Fusing object-level vector modality for explainable autonomous driving
Long Chen, Oleg Sinavski, Jan Hünermann, Alice Karnsund, Andrew James Willmott, Danny Birch, Daniel Maund, and Jamie Shotton. Driving with llms: Fusing object-level vector modality for explainable autonomous driving. In 2024 IEEE International Conference on Robotics and Automa...
2024
-
[18]
Optimizing autonomous driving for safety: A human-centric approach with llm-enhanced rlhf
Yuan Sun, Navid Salami Pargoo, Peter Jin, and Jorge Ortiz. Optimizing autonomous driving for safety: A human-centric approach with llm-enhanced rlhf. In Companion of the 2024 on ACM International Joint Conference on Pervasive and Ubiquitous Computing, pages 76–80, 2024
2024
-
[19]
O’Brien, Carrie J
Joon Sung Park, Joseph C. O’Brien, Carrie J. Cai, Meredith Ringel Morris, Percy Liang, and Michael S. Bernstein. Generative agents: Interactive simulacra of human behavior, April 01, 2023 2023
2023
-
[20]
Tenenbaum, and Igor Mordatch
Yilun Du, Shuang Li, Antonio Torralba, Joshua B. Tenenbaum, and Igor Mordatch. Improving factuality and reasoning in language models through multiagent debate. CoRR, abs/2305.14325, 2023
2023 arXiv
-
[21]
Metagpt: Meta programming for multi-agent collaborative framework, August 01, 2023 2023
Sirui Hong, Xiawu Zheng, Jonathan Chen, Yuheng Cheng, Jinlin Wang, Ceyao Zhang, Zili Wang, Steven Ka Shing Yau, Zijuan Lin, Liyang Zhou, Chenyu Ran, Lingfeng Xiao, and Chenglin Wu. Metagpt: Meta programming for multi-agent collaborative framework, August 01, 2023 2023
2023
-
[22]
Society of mind
Marvin Minsky. Society of mind. Simon and Schuster, 1988
1988
-
[23]
Examining the society of mind
Push Singh. Examining the society of mind. Comput. Artif. Intell., 22(6):521–543, 2003
2003
-
[24]
CAMEL: communicative agents for "mind" exploration of large language model society
Guohao Li, Hasan Hammoud, Hani Itani, Dmitrii Khizbullin, and Bernard Ghanem. CAMEL: communicative agents for "mind" exploration of large language model society. In NeurIPS, 2023
2023
-
[25]
Unleashing cognitive synergy in large language models: A task-solving agent through multi-persona self- collaboration, July 01, 2023 2023
Zhenhailong Wang, Shaoguang Mao, Wenshan Wu, Tao Ge, Furu Wei, and Heng Ji. Unleashing cognitive synergy in large language models: A task-solving agent through multi-persona self- collaboration, July 01, 2023 2023. work in progress
2023
-
[26]
Chawla, Olaf Wiest, and Xiangliang Zhang
Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V . Chawla, Olaf Wiest, and Xiangliang Zhang. Large language model based multi-agents: A survey of progress and challenges. CoRR, abs/2402.01680, 2024
2024 arXiv
-
[27]
Reasoning capacity in multi-agent systems: Limitations, challenges and human-centered solutions
Pouya Pezeshkpour, Eser Kandogan, Nikita Bhutani, Sajjadur Rahman, Tom Mitchell, and Estevam Hruschka. Reasoning capacity in multi-agent systems: Limitations, challenges and human-centered solutions. CoRR, abs/2402.01108, 2024
2024 arXiv
-
[28]
Cooperate or collapse: Emergence of sustainability behaviors in a society of llm agents
Giorgio Piatti, Zhijing Jin, Max Kleiman-Weiner, Bernhard Schölkopf, Mrinmaya Sachan, and Rada Mihalcea. Cooperate or collapse: Emergence of sustainability behaviors in a society of llm agents. arXiv preprint arXiv:2404.16698, 2024
2024 arXiv
-
[29]
Discovering causality for efficient coopera- tion in multi-agent environments
Rafael Pina, Varuna De Silva, and Corentin Artaud. Discovering causality for efficient coopera- tion in multi-agent environments. CoRR, abs/2306.11846, 2023
2023 arXiv
-
[30]
Cut the crap: An economical communication pipeline for llm-based multi-agent systems
Guibin Zhang, Yanwei Yue, Zhixun Li, Sukwon Yun, Guancheng Wan, Kun Wang, Dawei Cheng, Jeffrey Xu Yu, and Tianlong Chen. Cut the crap: An economical communication pipeline for llm-based multi-agent systems. arXiv preprint arXiv:2410.02506, 2024. 11
-
[31]
Masrouter: Learning to route llms for multi-agent systems
Yanwei Yue, Guibin Zhang, Boyang Liu, Guancheng Wan, Kun Wang, Dawei Cheng, and Yiyan Qi. Masrouter: Learning to route llms for multi-agent systems. arXiv preprint arXiv:2502.11133, 2025
2025 arXiv
-
[32]
Competeai: Understanding the competition behaviors in large language model-based agents
Qinlin Zhao, Jindong Wang, Yixuan Zhang, Yiqiao Jin, Kaijie Zhu, Hao Chen, and Xing Xie. Competeai: Understanding the competition behaviors in large language model-based agents. arXiv preprint arXiv:2310.17512, 2023
2023 arXiv
-
[33]
Encouraging divergent thinking in large language models through multi-agent debate
Tian Liang, Zhiwei He, Wenxiang Jiao, Xing Wang, Yan Wang, Rui Wang, Yujiu Yang, Zhaopeng Tu, and Shuming Shi. Encouraging divergent thinking in large language models through multi-agent debate. CoRR, abs/2305.19118, 2023
2023 arXiv
-
[34]
Battleagentbench: A benchmark for evaluating cooperation and competition capabilities of language models in multi-agent systems
Wei Wang, Dan Zhang, Tao Feng, Boyan Wang, and Jie Tang. Battleagentbench: A benchmark for evaluating cooperation and competition capabilities of language models in multi-agent systems. arXiv preprint arXiv:2408.15971, 2024
2024 arXiv
-
[35]
Progressive-hint prompt- ing improves reasoning in large language models, April 01, 2023 2023
Chuanyang Zheng, Zhengying Liu, Enze Xie, Zhenguo Li, and Yu Li. Progressive-hint prompt- ing improves reasoning in large language models, April 01, 2023 2023. Tech Report
2023
-
[36]
Memorybank: Enhancing large language models with long-term memory
Wanjun Zhong, Lianghong Guo, Qiqi Gao, He Ye, and Yanlin Wang. Memorybank: Enhancing large language models with long-term memory. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 19724–19731, 2024
2024
-
[37]
Memgpt: Towards llms as operating systems
Charles Packer, Vivian Fang, Shishir_G Patil, Kevin Lin, Sarah Wooders, and Joseph_E Gonza- lez. Memgpt: Towards llms as operating systems. 2023
2023
-
[38]
Mem- llm: Finetuning llms to use an explicit read-write memory
Ali Modarressi, Abdullatif Köksal, Ayyoob Imani, Mohsen Fayyaz, and Hinrich Schütze. Mem- llm: Finetuning llms to use an explicit read-write memory. arXiv preprint arXiv:2404.11672, 2024
2024 arXiv
-
[39]
A survey on the memory mechanism of large language model based agents
Zeyu Zhang, Xiaohe Bo, Chen Ma, Rui Li, Xu Chen, Quanyu Dai, Jieming Zhu, Zhenhua Dong, and Ji-Rong Wen. A survey on the memory mechanism of large language model based agents. arXiv preprint arXiv:2404.13501, 2024
2024 arXiv
-
[40]
Chatdb: Augmenting llms with databases as their symbolic memory
Chenxu Hu, Jie Fu, Chenzhuang Du, Simian Luo, Junbo Zhao, and Hang Zhao. Chatdb: Augmenting llms with databases as their symbolic memory. arXiv preprint arXiv:2306.03901, 2023
2023 arXiv
-
[41]
Memochat: Tuning llms to use memos for consistent long-range open-domain conversation
Junru Lu, Siyu An, Mingbao Lin, Gabriele Pergola, Yulan He, Di Yin, Xing Sun, and Yunsheng Wu. Memochat: Tuning llms to use memos for consistent long-range open-domain conversation. arXiv preprint arXiv:2308.08239, 2023
2023 arXiv
-
[42]
Recmind: Large language model powered agent for recommendation
Yancheng Wang, Ziyan Jiang, Zheng Chen, Fan Yang, Yingxue Zhou, Eunah Cho, Xing Fan, Xiaojiang Huang, Yanbin Lu, and Yingzhen Yang. Recmind: Large language model powered agent for recommendation. arXiv preprint arXiv:2308.14296, 2023
2023 arXiv
-
[43]
Expel: Llm agents are experiential learners
Andrew Zhao, Daniel Huang, Quentin Xu, Matthieu Lin, Yong-Jin Liu, and Gao Huang. Expel: Llm agents are experiential learners. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 19632–19642, 2024
2024
-
[44]
Metaagents: Simulating interactions of human behaviors for llm-based task-oriented coordination via collaborative generative agents
Yuan Li, Yixuan Zhang, and Lichao Sun. Metaagents: Simulating interactions of human behaviors for llm-based task-oriented coordination via collaborative generative agents. arXiv preprint arXiv:2310.06500, 2023
2023 arXiv
-
[45]
S3: Social-network simulation system with large language model-empowered agents
Chen Gao, Xiaochong Lan, Zhihong Lu, Jinzhu Mao, Jinghua Piao, Huandong Wang, Depeng Jin, and Yong Li. S3: Social-network simulation system with large language model-empowered agents. arXiv preprint arXiv:2307.14984, 2023
2023 arXiv
-
[46]
Communicative agents for software development, July 01, 2023 2023
Chen Qian, Xin Cong, Cheng Yang, Weize Chen, Yusheng Su, Juyuan Xu, Zhiyuan Liu, and Maosong Sun. Communicative agents for software development, July 01, 2023 2023. 25 pages, 9 figures, 2 tables
2023
-
[47]
Scaling large-language-model-based multi-agent collab- oration
Chen Qian, Zihao Xie, Yifei Wang, Wei Liu, Yufan Dang, Zhuoyun Du, Weize Chen, Cheng Yang, Zhiyuan Liu, and Maosong Sun. Scaling large-language-model-based multi-agent collab- oration. arXiv preprint arXiv:2406.07155, 2024. 12
2024 arXiv
-
[48]
Gptswarm: Language agents as optimizable graphs
Mingchen Zhuge, Wenyi Wang, Louis Kirsch, Francesco Faccio, Dmitrii Khizbullin, and Jürgen Schmidhuber. Gptswarm: Language agents as optimizable graphs. In Forty-first International Conference on Machine Learning, 2024
2024
-
[49]
Automated design of agentic systems
Shengran Hu, Cong Lu, and Jeff Clune. Automated design of agentic systems. arXiv preprint arXiv:2408.08435, 2024
2024 arXiv
-
[50]
AFlow: Automating Agentic Workflow Generation, October 2024
Jiayi Zhang, Jinyu Xiang, Zhaoyang Yu, Fengwei Teng, Xionghui Chen, Jiaqi Chen, Mingchen Zhuge, Xin Cheng, Sirui Hong, Jinlin Wang, Bingnan Zheng, Bang Liu, Yuyu Luo, and Chenglin Wu. AFlow: Automating Agentic Workflow Generation, October 2024. arXiv:2410.10762
2024 arXiv
-
[51]
Multi-agent architecture search via agentic supernet
Guibin Zhang, Luyang Niu, Junfeng Fang, Kun Wang, Lei Bai, and Xiang Wang. Multi-agent architecture search via agentic supernet. arXiv preprint arXiv:2502.04180, 2025
2025 arXiv
-
[52]
Exchange-of-thought: Enhancing large language model capabilities through cross-model communication
Zhangyue Yin, Qiushi Sun, Cheng Chang, Qipeng Guo, Junqi Dai, Xuan-Jing Huang, and Xipeng Qiu. Exchange-of-thought: Enhancing large language model capabilities through cross-model communication. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Pro...
2023
-
[53]
A survey on large language model based autonomous agents
Lei Wang, Chen Ma, Xueyang Feng, Zeyu Zhang, Hao Yang, Jingsen Zhang, Zhiyuan Chen, Jiakai Tang, Xu Chen, Yankai Lin, Wayne Xin Zhao, Zhewei Wei, and Ji-Rong Wen. A survey on large language model based autonomous agents. Front. Comput. Sci., 18, 2024
2024
-
[54]
The rise and potential of large language model based agents: A survey
Zhiheng Xi, Wenxiang Chen, Xin Guo, Wei He, Yiwen Ding, Boyang Hong, Ming Zhang, Junzhe Wang, Senjie Jin, Enyu Zhou, Rui Zheng, Xiaoran Fan, Xiao Wang, Limao Xiong, Yuhao Zhou, Weiran Wang, Changhao Jiang, Yicheng Zou, Xiangyang Liu, Zhangyue Yin, Shihan Dou, Rongxiang Weng, W...
2023 arXiv
-
[55]
Large language models empowered agent-based modeling and simulation: A survey and perspectives
Chen Gao, Xiaochong Lan, Nian Li, Yuan Yuan, Jingtao Ding, Zhilun Zhou, Fengli Xu, and Yong Li. Large language models empowered agent-based modeling and simulation: A survey and perspectives. CoRR, abs/2312.11970, 2023
2023 arXiv
-
[56]
A survey on llm-based multi-agent systems: workflow, infrastructure, and challenges
Xinyi Li, Sai Wang, Siqi Zeng, Yu Wu, and Yi Yang. A survey on llm-based multi-agent systems: workflow, infrastructure, and challenges. Vicinagearth, 1(1):9, 2024
2024
-
[57]
Synapse: Trajectory-as-exemplar prompting with memory for computer control
Longtao Zheng, Rundong Wang, Xinrun Wang, and Bo An. Synapse: Trajectory-as-exemplar prompting with memory for computer control. arXiv preprint arXiv:2306.07863, 2023
2023 arXiv
-
[58]
Ghost in the minecraft: Generally capable agents for open-world environments via large language models with text-based knowledge and memory
Xizhou Zhu, Yuntao Chen, Hao Tian, Chenxin Tao, Weijie Su, Chenyu Yang, Gao Huang, Bin Li, Lewei Lu, Xiaogang Wang, et al. Ghost in the minecraft: Generally capable agents for open-world environments via large language models with text-based knowledge and memory. arXiv preprin...
2023 arXiv
-
[59]
Chemagent: Self-updating library in large language models improves chemical reasoning
Xiangru Tang, Tianyu Hu, Muyang Ye, Yanjun Shao, Xunjian Yin, Siru Ouyang, Wangchunshu Zhou, Pan Lu, Zhuosheng Zhang, Yilun Zhao, et al. Chemagent: Self-updating library in large language models improves chemical reasoning. arXiv preprint arXiv:2501.06590, 2025
2025 arXiv
-
[60]
Reflexion: an autonomous agent with dynamic memory and self-reflection
Noah Shinn, Beck Labash, and Ashwin Gopinath. Reflexion: an autonomous agent with dynamic memory and self-reflection. arXiv preprint, abs/2303.11366, 2023
2023 arXiv
-
[61]
A-mem: Agentic memory for llm agents
Wujiang Xu, Kai Mei, Hang Gao, Juntao Tan, Zujie Liang, and Yongfeng Zhang. A-mem: Agentic memory for llm agents. arXiv preprint arXiv:2502.12110, 2025
2025 arXiv
-
[62]
Mem0: Building production-ready ai agents with scalable long-term memory
Prateek Chhikara, Dev Khant, Saket Aryan, Taranjeet Singh, and Deshraj Yadav. Mem0: Building production-ready ai agents with scalable long-term memory. arXiv preprint arXiv:2504.19413, 2025
2025 arXiv
-
[63]
Meminsight: Autonomous memory augmentation for llm agents
Rana Salama, Jason Cai, Michelle Yuan, Anna Currey, Monica Sunkara, Yi Zhang, and Yassine Benajiba. Meminsight: Autonomous memory augmentation for llm agents. arXiv preprint arXiv:2503.21760, 2025. 13
2025 arXiv
-
[64]
Mixture-of-agents enhances large language model capabilities
Junlin Wang, Jue Wang, Ben Athiwaratkun, Ce Zhang, and James Zou. Mixture-of-agents enhances large language model capabilities. arXiv preprint arXiv:2406.04692, 2024
2024 arXiv
-
[65]
Symbolic learning enables self-evolving agents
Wangchunshu Zhou, Yixin Ou, Shengwei Ding, Long Li, Jialong Wu, Tiannan Wang, Jiamin Chen, Shuai Wang, Xiaohua Xu, Ningyu Zhang, et al. Symbolic learning enables self-evolving agents. arXiv preprint arXiv:2406.18532, 2024
2024 arXiv
-
[66]
Self-evolving agents with reflective and memory-augmented abilities
Xuechen Liang, Meiling Tao, Yinghui Xia, Tianyu Shi, Jun Wang, and JingSong Yang. Self-evolving agents with reflective and memory-augmented abilities. arXiv preprint arXiv:2409.00872, 2024
2024 arXiv
-
[67]
Agentverse: Facilitating multi-agent collaboration and exploring emergent behaviors in agents, 2023
Weize Chen, Yusheng Su, Jingwei Zuo, Cheng Yang, Chenfei Yuan, Chen Qian, Chi-Min Chan, Yujia Qin, Yaxi Lu, Ruobing Xie, Zhiyuan Liu, Maosong Sun, and Jie Zhou. Agentverse: Facilitating multi-agent collaboration and exploring emergent behaviors in agents, 2023
2023
-
[68]
Self-evolving multi-agent collaboration networks for software development
Yue Hu, Yuzhu Cai, Yaxin Du, Xinyu Zhu, Xiangrui Liu, Zijie Yu, Yuchen Hou, Shuo Tang, and Siheng Chen. Self-evolving multi-agent collaboration networks for software development. arXiv preprint arXiv:2410.16946, 2024
2024 arXiv
-
[69]
G-designer: Architecting multi-agent communication topologies via graph neural networks
Guibin Zhang, Yanwei Yue, Xiangguo Sun, Guancheng Wan, Miao Yu, Junfeng Fang, Kun Wang, Tianlong Chen, and Dawei Cheng. G-designer: Architecting multi-agent communication topologies via graph neural networks. arXiv preprint arXiv:2410.11782, 2024
-
[70]
Evoa- gent: Towards automatic multi-agent generation via evolutionary algorithms
Siyu Yuan, Kaitao Song, Jiangjie Chen, Xu Tan, Dongsheng Li, and Deqing Yang. Evoa- gent: Towards automatic multi-agent generation via evolutionary algorithms. arXiv preprint arXiv:2406.14228, 2024
2024 arXiv
-
[71]
Evoflow: Evolving diverse agentic workflows on the fly
Guibin Zhang, Kaijie Chen, Guancheng Wan, Heng Chang, Hong Cheng, Kun Wang, Shuyue Hu, and Lei Bai. Evoflow: Evolving diverse agentic workflows on the fly. arXiv preprint arXiv:2502.07373, 2025
2025 arXiv
-
[72]
Dynamic llm-agent network: An llm-agent collaboration framework with agent team optimization
Zijun Liu, Yanzhe Zhang, Peng Li, Yang Liu, and Diyi Yang. Dynamic llm-agent network: An llm-agent collaboration framework with agent team optimization. CoRR, abs/2310.02170, 2023
2023 arXiv
-
[73]
Microsoft academic graph: When experts are not enough
Kuansan Wang, Zhihong Shen, Chiyuan Huang, Chieh-Han Wu, Yuxiao Dong, and Anshul Kanakia. Microsoft academic graph: When experts are not enough. Quantitative Science Studies, 1(1):396–413, 2020
2020
-
[74]
Sirius: Self-improving multi- agent systems via bootstrapped reasoning
Wanjia Zhao, Mert Yuksekgonul, Shirley Wu, and James Zou. Sirius: Self-improving multi- agent systems via bootstrapped reasoning. arXiv preprint arXiv:2502.04780, 2025
2025 arXiv
-
[75]
Reso: A reward-driven self- organizing llm-based multi-agent system for reasoning tasks
Heng Zhou, Hejia Geng, Xiangyuan Xue, Zhenfei Yin, and Lei Bai. Reso: A reward-driven self- organizing llm-based multi-agent system for reasoning tasks. arXiv preprint arXiv:2503.02390, 2025
2025 arXiv
-
[76]
Hotpotqa: A dataset for diverse, explainable multi-hop question answering
Zhilin Yang, Peng Qi, Saizheng Zhang, Yoshua Bengio, William W Cohen, Ruslan Salakhut- dinov, and Christopher D Manning. Hotpotqa: A dataset for diverse, explainable multi-hop question answering. arXiv preprint arXiv:1809.09600, 2018
2018 arXiv
-
[77]
Fever: a large-scale dataset for fact extraction and verification
James Thorne, Andreas Vlachos, Christos Christodoulopoulos, and Arpit Mittal. Fever: a large-scale dataset for fact extraction and verification. arXiv preprint arXiv:1803.05355, 2018
2018 arXiv
-
[78]
Alfworld: Aligning text and embodied environments for interactive learning
Mohit Shridhar, Xingdi Yuan, Marc-Alexandre Côté, Yonatan Bisk, Adam Trischler, and Matthew Hausknecht. Alfworld: Aligning text and embodied environments for interactive learning. arXiv preprint arXiv:2010.03768, 2020
2010 arXiv
-
[79]
Scienceworld: Is your agent smarter than a 5th grader? arXiv preprint arXiv:2203.07540, 2022
Ruoyao Wang, Peter Jansen, Marc-Alexandre Côté, and Prithviraj Ammanabrolu. Scienceworld: Is your agent smarter than a 5th grader? arXiv preprint arXiv:2203.07540, 2022
2022 arXiv
-
[80]
Agentboard: An analytical evaluation board of multi-turn llm agents
Chang Ma, Junlei Zhang, Zhihao Zhu, Cheng Yang, Yujiu Yang, Yaohui Jin, Zhenzhong Lan, Lingpeng Kong, and Junxian He. Agentboard: An analytical evaluation board of multi-turn llm agents. arXiv preprint arXiv:2401.13178, 2024. 14
2024 arXiv
-
[81]
You are a smart agent designed to solve problems
Wenhui Wang, Furu Wei, Li Dong, Hangbo Bao, Nan Yang, and Ming Zhou. Minilm: Deep self-attention distillation for task-agnostic compression of pre-trained transformers. Advances in Neural Information Processing Systems, 33:5776–5788, 2020. Impact Statement G-Memory introduces ...
2020
-
[82]
"" l e a r n _ l e s s o n s _ u s e r _ p r o m p t _ c o m p a r e =
Insight 3 ... """ l e a r n _ l e s s o n s _ u s e r _ p r o m p t _ c o m p a r e = """ ## S u c c e s s f u l t r a j e c t o r y { t r u e _ t r a j } ## Failed t r a j e c t o r y ### t r a j e c t o r y { f a l s e _ t r a j } Your output : """ l e a r n _ l e s s o n s ...
1920
-
[83]
"" l e a r n _ l e s s o n s _ u s e r _ p r o m p t _ a l l _ s u c c =
Insight 3 ... """ l e a r n _ l e s s o n s _ u s e r _ p r o m p t _ a l l _ s u c c = """ ## S u c c e s s f u l t r a j e c t o r y s { t r u e _ t r a j s } 22 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1920 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 4...
1920
-
[84]
"" m e r g e _ r u l e s _ u s e r _ p r o m p t =
Insight 3 ... """ m e r g e _ r u l e s _ u s e r _ p r o m p t = """ ## Here are the current i ns igh ts that need to be merged : { c u r r e n t _ r u l e s } ## Please c o n s o l i d a t e and rewrite them into ** no more than { l i m i t e d _ n u m b e r } refined ins ig...
-
[85]
"" p r o j e c t _ i n s i g h t s _ u s e r _ p r o m p t : str =
Insight 3 ... """ p r o j e c t _ i n s i g h t s _ u s e r _ p r o m p t : str = """ ### T r a j e c t o r y { t r a j e c t o r y } ### Agent ’ s Role : { role } ### General In si gh ts : { ins ig ht s } ### Your Output ( P e r s o n a l i z e d In si ght s for This Role ) :...
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.