Pith. sign in

REVIEW 2 major objections 4 minor 42 references

Learn as Individuals, Evolve as a Team: Multi-agent LLMs Adaptation in Embodied Environments

T0 review · 2 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Multi-agent LLM planners adapt best when each agent learns a cost-prediction utility function and the team evolves a shared communication knowledge list, outperforming established baselines on two embodied benchmarks.

desk verdict A sensible multi-agent LLM adaptation framework whose headline gains rest on an unverified train/eval split that the paper's own numbers make look suspicious. read the letter →

arxiv 2506.07232 v1 pith:MOT5O3GY submitted 2025-06-08 cs.MA cs.AIcs.LG

Xinran Li , Chenjia Bai , Zijian Li , Jiakun Zheng , Ting Xiao , Jun Zhang This is my paper
classification cs.MAcs.AIcs.LG
keywords multi-agentplanningLLMagentsembodiedenvironmentsutilityfunctiontest-timeadaptationcooperativecommunicationpromptevolutionDec-POMDP
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes LIET—'Learn as Individuals, Evolve as a Team'—a semi-centralized recipe for making large-language-model agents plan together in embodied household environments. Its central claim is that LLM agents adapt better when each agent first learns a cost-prediction utility function from exploratory data and the team then evolves a shared cooperation-knowledge list through receiver-side reflection during test time. On the C-WAH and TDW-MAT benchmarks, with both LLaMA-3.1-70B and GPT-4o as backbones, LIET outperforms established decentralized and centralized baselines. If the claim is right, the bottleneck for multi-agent LLM planning in embodied settings is adaptation through experience rather than zero-shot prompting alone.

What carries the argument

The two mechanisms that carry LIET are the utility function and the shared cooperation knowledge list. The utility function is a parameter-efficiently fine-tuned LLM with an MLP value head that predicts per-action step costs from text descriptions; it supplies the individual learning that lets a planner prefer cheap, feasible actions. The cooperation knowledge list is a shared, evolving prompt of message-construction tips, maintained by a receiver-side reflector that evaluates incoming messages and merges insights back into the list; it supplies the team-level adaptation that makes communication increasingly specific and useful. The framework wraps both around a decentralized planner that broadcasts messages only at critical timesteps, which is what makes the design semi-centralized.

What would settle it

Re-run the C-WAH and TDW-MAT evaluations with a training/evaluation split guaranteed disjoint—for example, hold out entire room layouts or task types from utility-function training—and compare full LIET against its 'without individual learning' ablation; if the gap shrinks to near zero, the utility gain is memorization of evaluation episodes rather than generalization.

Watch

Extended reading notes

Core claim

LIET's central claim is that two layers of adaptation close the gap between off-the-shelf LLM planners and competent embodied teammates. In the individual layer, each agent queries a utility function $f(\ell_{o_i}, \ell_a)$ that returns the estimated number of environment steps needed to execute an action from the current observation; this function is a 1B-parameter LLM with an MLP value head, fine-tuned by mean-squared-error regression on exploratory episodes collected by agents themselves, so no human cost labels are needed. In the team layer, agents maintain a shared cooperation knowledge list—a growing set of communication tips—which is re-prompted into every message-generation call and is updated after each exchange by the receiver, who reflects on whether and how the incoming message helped planning. The paper reports that on C-WAH and TDW-MAT, with LLaMA-3.1-70B and GPT-4o as backbones, LIET achieves lower average task-completion steps and higher transport rates than the compared baselines, and its ablations attribute the gain to both layers rather than to one component.

Load-bearing premise

The load-bearing premise is that the exploratory episodes teach a transferable cost function: the paper reports collecting 10 C-WAH and 24 TDW-MAT episodes—the same counts as the evaluation sets—and does not show the training and evaluation episodes are disjoint, so the individual-learning improvement could in part reflect memorized costs.

Editorial extensions

If this is right

  • LIET's individual utility function can be trained from self-collected exploration data, so adapting a planner to a new embodied environment does not require human expert labels for action costs.
  • The team-level knowledge list gives communication a test-time memory: messages become more specific as episodes progress, down to object-level task division, as the paper's visualizations show.
  • The utility function trained only on two-agent episodes transfers to three- and four-agent settings without retraining, which the paper presents as evidence of generalization to larger teams.
  • The C-WAH ablations show that both layers contribute: removing the learned utility function, replacing cost estimates with prompted guesses, or freezing the communication prompt all hurt performance.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct test the paper leaves implicit: publish the exploratory episodes and re-run with a guaranteed disjoint train/eval split; this would show whether the individual-learning gain reflects generalization or memorization of the fixed evaluation episodes.
  • The goal-agnostic cost predictor could be reused across tasks within the same simulator; reporting per-task cost-prediction error would indicate how much of the C-WAH gain is environment knowledge versus task-specific tuning.
  • The same two-layer pattern—a learned scalar cost/utility head on a small LLM, plus a shared prompt list updated by receiver-side reflection—could be dropped into other decentralized multi-agent settings such as warehouse or traffic coordination, since neither component depends on household chore structure.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. The paper proposes LIET (Learn as Individuals, Evolve as a Team), a semi-centralized framework for adapting LLM-based planners to multi-agent embodied environments. At the individual level, agents use a utility function, implemented by fine-tuning a small LLM with LoRA and a value head, to estimate action costs from exploratory data. At the team level, agents maintain a shared cooperation knowledge list that is iteratively updated through receiver-side reflection, and this list guides message generation during test time. The method is evaluated on Communicative Watch-And-Help (C-WAH) and ThreeDWorld Multi-Agent Transport (TDW-MAT) with LLaMA-3.1-70B and GPT-4o backbones, compared against CoELA, ProAgent, and RoCo. The paper reports improved task-completion steps and transport rates, ablation studies showing the utility function and evolving communication both contribute, and additional experiments with 3 and 4 agents.

Significance. If the reported results are sound, the paper makes a useful contribution to multi-agent LLM planning by combining two relatively lightweight adaptation mechanisms: a learned cost estimator for individual decision-making and an evolving communication prompt for team coordination. The framework is simple, modular, and potentially applicable to other embodied multi-agent settings, and the visualization and ablation studies help explain why each component matters. However, the central empirical claims currently lack statistical support, and there is a serious unresolved question about whether the utility function's training data overlaps with the evaluation episodes. Because both issues directly affect the validity of the headline comparisons, the significance of the contribution cannot be assessed from the manuscript as written.

major comments (2)
  1. [Section 4.1 vs. Section A.1, Table 3] The paper must clarify whether the exploratory dataset used to fine-tune the utility function is disjoint from the evaluation episodes. Section 4.1 states that the C-WAH evaluation set consists of 10 episodes and the TDW-MAT evaluation set includes 24 episodes. Section A.1 then states that the exploratory dataset was collected from 10 C-WAH episodes and 24 TDW-MAT episodes, with 80% used as the training set. These numbers match exactly, and no sentence in the paper says that the exploratory episodes come from different task instances, object layouts, seeds, or floorplans. Since the utility function is trained on (observation, action, cost) pairs and then queried during evaluation, any overlap would let it memorize episode-specific object locations and action costs rather than learn transferable environment knowledge. Table 3 shows that removing the individual utility module degrades C-WAH performance from 40.3 to 52.7 steps, so this potential leakage directly inflates the reported contribution of the 'learn as individuals' component. The authors need to either demonstrate disjointness explicitly or re-run the evaluation with held-out episodes before the headline claim can be accepted.
  2. [Tables 1, 2, 3, and Section 4.1] The reported results lack any measure of statistical uncertainty. The implementation details state that LLM inference uses temperature = 0.7, so the planning process is stochastic, yet Tables 1 and 2 report only single average values for each method without standard deviations, confidence intervals, or significance tests. The abstract's claim that LIET 'significantly outperforms' baselines is therefore not supported by the evidence presented. The same issue applies to the ablation results in Table 3 and the multi-agent results in Table 4. At minimum, the paper should report means and standard deviations over multiple independent runs (with distinct random seeds) and perform a standard significance test for the main comparisons against CoELA, ProAgent, and RoCo.
minor comments (4)
  1. [Section 3.3] There is a typo in the message generator prompting template: 'placehoder' should be 'placeholder'.
  2. [Section 3] The heading 'Learn as Individuals, Envolve as a Team' contains a typo; 'Envolve' should be 'Evolve'.
  3. [Section 4.1] The baseline description says 'A-start-based low-level planner'; this should be 'A*' or 'A-star'.
  4. [Section A.1] The paper states that code will be released after review, but it does not report the exact split of the exploratory dataset into training and validation, nor does it give the number of description-cost pairs per episode. Providing these details would improve reproducibility and help address the data-composition concern.

Circularity Check

1 steps flagged · score 6.0 of 10

The utility function is finetuned on exploratory datasets whose episode counts match the evaluation sets exactly (10 C-WAH, 24 TDW-MAT), with no stated disjointness; if those episodes overlap, the individual-learning gain reduces to memorized cost fits rather than prediction.

  1. fitted input called prediction [Section 3.2 Eq. (1), Section 4.1, Section A.1]
    ""For the exploratory dataset, we collected 10 episodes for C-WAH benchmark and 24 episodes for TDW-MAT benchmark. We used 80% of this data as the training set to finetune a LLaMA 3.2-1B model with LoRA adapters and MLP layers as the value head." ... "The evaluation set consists of 10 episodes spanning five different tasks, each with a horizon of 250 timesteps." ... "The evaluation set includes 24 episodes divided into 'food' and 'stuff' tasks, with each episode spanning up to 3000 time frames.""

    Eq. (1) fits f by MSE on the exploratory dataset; the paper's headline result is measured on exactly 10 C-WAH and 24 TDW-MAT evaluation episodes. Section A.1 never states that the exploratory episodes are disjoint from the evaluation episodes, and the counts match exactly. If the exploratory episodes are the evaluation episodes, the 80/20 split still puts 8 of 10 C-WAH evaluation episodes into the utility training set; for TDW-MAT, Sec. A.2.1 defines separate training/test floorplans, but A.1 does not say the 24 exploratory episodes come from the training floorplans. The test-time cost c=f(ℓ_oi,ℓ_a) is then a fitted value on evaluation instances, so the Table 3 gain (40.3 vs 52.7 without individual learning) is memorization presented as prediction.

full rationale

The only load-bearing circularity concern is the utility-finettuning step. The paper's derivation chain is: collect exploratory data, fit a cost-prediction utility with MSE, query it at test time, and attribute part of LIET's improvement to this individual-learning module. The paper's own numbers allow the exploratory dataset to coincide with the evaluation set: both are 10 C-WAH episodes and 24 TDW-MAT episodes, and no sentence establishes disjointness. If the sets overlap, the utility's outputs are fitted values for the evaluation instances, and the ablation gain is not evidence of generalization. I score this 6 rather than 8 because the circularity is conditional on an unstated data-overlap assumption, and because the team-evolving communication module and the comparisons against external baselines are not themselves circular. A single explicit statement that exploratory episodes use different layouts/seeds/task instances than evaluation episodes would remove the issue; as written, the prediction claim is not protected against reduction to a fit.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the assumption that the utility function generalizes from its training data to the evaluation tasks, and that the LLM-based message generation and reflection loop improve coordination. The paper provides no formal guarantee and leaves the train or eval overlap unspecified. No new physical entities are introduced.

assumptions (3)
  • domain assumption The exploratory datasets used to finetune the utility function are representative of, and disjoint from, the evaluation episodes.
    Section A.1 says the datasets are 10 episodes for C-WAH and 24 for TDW-MAT, matching the evaluation set sizes, but does not state whether they are the same episodes. The validity of the performance comparison depends on this.
  • domain assumption The LLM planner reliably incorporates the utility estimates and the knowledge list into its decisions.
    The method assumes that prompting the LLM with action costs and communication hints changes behavior in the intended direction; this is supported only by ablation, not by direct analysis.
  • domain assumption The reward and action costs used for the utility function are accurate measures of environment dynamics.
    The utility function is trained to predict the number of steps, so the ground-truth labels from exploratory rollouts must faithfully represent execution costs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learn as Individuals, Evolve as a Team: Multi-agent LLMs Adaptation in Embodied Environments." pith.science (2026). https://pith.science/paper/MOT5O3GY

@misc{pith2026250607232,
  author       = {Pith},
  title        = {Pith review of: Learn as Individuals, Evolve as a Team: Multi-agent LLMs Adaptation in Embodied Environments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MOT5O3GY}},
  note         = {Machine review of arXiv:2506.07232}
}
read the original abstract

Large language models (LLMs) possess extensive knowledge bases and strong reasoning capabilities, making them promising tools for complex, multi-agent planning in embodied environments. However, despite LLMs' advanced abilities and the sophisticated modular design of agentic methods, existing LLM-based planning algorithms remain limited by weak adaptation capabilities to multi-agent embodied scenarios. We address this limitation by introducing a framework that enables LLM agents to learn and evolve both before and during test time, equipping them with environment-relevant knowledge for better planning and enhanced communication for improved cooperation. Inspired by centralized training with decentralized execution in multi-agent reinforcement learning, we propose a \textit{Learn as Individuals, Evolve as a Team (LIET)} paradigm for multi-agent LLMs adaptation. At the individual level, LLM agents learn a local utility function from exploratory datasets to better comprehend the embodied environment, which is then queried during test time to support informed decision-making. At the team level, LLM agents collaboratively and iteratively maintain and update a shared cooperation knowledge list based on new experiences, using it to guide more effective communication. By combining individual learning with team evolution, LIET enables comprehensive and flexible adaptation for LLM agents. Our experiments on Communicative Watch-And-Help and ThreeD-World Multi-Agent Transport benchmarks demonstrate that LIET, instantiated with both LLaMA and GPT-4o, outperforms existing baselines and exhibits strong cooperative planning abilities.

Figures

Figures reproduced from arXiv: 2506.07232 by the authors.

Figure 1
Figure 1. The overall framework of LIET. It adopts a semi-centralized decision making scheme with [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. The scheme for LIET communication. In LIET, communication happens at critical timesteps [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Snapshots of an example episode from the C-WAH benchmark with the task: “Find and [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Snapshots of an example episode from the C-WAH benchmark with the task: “Find and [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

42 extracted references · 40 canonical work pages

  1. [1]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al . 2023. Gpt-4 technical report. Technical Report. OpenAI

    work page 2023

  2. [2]

    Anthony Brohan, Yevgen Chebotar, Chelsea Finn, Karol Hausman, Alexander Herzog, Daniel Ho, Julian Ibarz, Alex Irpan, Eric Jang, Ryan Julian, et al . 2023. Do as i can, not as i say: Grounding language in robotic affordances. InProceedings of the Conference on Robot Learning. PMLR, 287–318

    work page 2023

  3. [3]

    Jie Cai, Donghun Kim, Rita Jaramillo, James E Braun, and Jianghai Hu. 2016. A general multi-agent control approach for building energy system optimization.Energy and Buildings 127 (2016), 337–351

    work page 2016

  4. [4]

    Yongchao Chen, Jacob Arkin, Yang Zhang, Nicholas Roy, and Chuchu Fan. 2024. Scalable multi-robot collaboration with large language models: Centralized or decentralized systems?. In 2024 IEEE International Conference on Robotics and Automation. IEEE, 4311–4317

    work page 2024

  5. [5]

    Tianshu Chu, Jie Wang, Lara Codecà, and Zhaojian Li. 2019. Multi-agent deep reinforcement learning for large-scale traffic signal control.IEEE transactions on intelligent transportation systems21, 3 (2019), 1086–1095

    work page 2019

  6. [6]

    Chuang Gan, Siyuan Zhou, Jeremy Schwartz, Seth Alter, Abhishek Bhandwaldar, Dan Gut- freund, Daniel LK Yamins, James J DiCarlo, Josh McDermott, Antonio Torralba, et al. 2022. The threedworld transport challenge: A visually guided task-and-motion planning benchmark towards physically realistic embodied ai. In2022 International Conference on Robotics and Au...

    work page 2022

  7. [7]

    Mihalis Giannakis and Michalis Louis. 2011. A multi-agent based framework for supply chain risk management.Journal of Purchasing and Supply Management17, 1 (2011), 23–31

    work page 2011

  8. [8]

    Ran Gong, Qiuyuan Huang, Xiaojian Ma, Yusuke Noda, Zane Durante, Zilong Zheng, Demetri Terzopoulos, Li Fei-Fei, Jianfeng Gao, and Hoi V o. 2024. MindAgent: Emergent Gaming Interaction. InFindings of the Association for Computational Linguistics. 3154–3183

    work page 2024

Show all 42 references
  1. [9]

    Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, et al . 2024. The llama 3 herd of models. Technical Report. Meta AI

    2024

  2. [10]

    2025.Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al. 2025.Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. Technical Report. DeepSeek-AI

    2025

  3. [11]

    Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V Chawla, Olaf Wiest, and Xiangliang Zhang. 2024. Large language model based multi-agents: A survey of progress and challenges. InProceedings of the 33rd International Joint Conference on Artificial Intelligence

    2024

  4. [12]

    Edward J Hu, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, Weizhu Chen, et al. 2022. LoRA: Low-Rank Adaptation of Large Language Models. InProceedings of the 10th International Conference on Learning Representations

    2022

  5. [13]

    Chenguang Huang, Oier Mees, Andy Zeng, and Wolfram Burgard. 2023. Visual language maps for robot navigation. In2023 IEEE International Conference on Robotics and Automation. IEEE, 10608–10615. 10

    2023

  6. [14]

    Wenlong Huang, Fei Xia, Dhruv Shah, Danny Driess, Andy Zeng, Yao Lu, Pete Florence, Igor Mordatch, Sergey Levine, Karol Hausman, et al. 2024. Grounded decoding: Guiding text generation with grounded models for embodied agents. InAdvances in Neural Information Processing System...

    2024

  7. [15]

    Wenlong Huang, Fei Xia, Ted Xiao, Harris Chan, Jacky Liang, Pete Florence, Andy Zeng, Jonathan Tompson, Igor Mordatch, Yevgen Chebotar, et al. 2023. Inner Monologue: Embodied Reasoning through Planning with Language Models. InProceedings of the Conference on Robot Learning. PM...

    2023

  8. [16]

    2024.Gpt-4o system card

    Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, et al. 2024.Gpt-4o system card. Technical Report. OpenAI

    2024

  9. [17]

    Shyam Sundar Kannan, Vishnunandan LN Venkatesh, and Byung-Cheol Min. 2024. Smart- llm: Smart multi-agent robot task planning using large language models. In2024 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 12140–12147

    2024

  10. [18]

    J-H Lee and C-O Kim. 2008. Multi-agent systems applications in manufacturing systems and supply chain management: a review paper.International Journal of Production Research46, 1 (2008), 233–265

    2008

  11. [19]

    Huao Li, Yu Chong, Simon Stepputtis, Joseph P Campbell, Dana Hughes, Charles Lewis, and Katia Sycara. 2023. Theory of Mind for Multi-Agent Collaboration via Large Language Models. InProceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 180–192

    2023

  12. [20]

    Bill Yuchen Lin, Yicheng Fu, Karina Yang, Faeze Brahman, Shiyu Huang, Chandra Bhagavatula, Prithviraj Ammanabrolu, Yejin Choi, and Xiang Ren. 2024. Swiftsage: A generative agent with fast and slow thinking for complex interactive tasks.Advances in Neural Information Processing...

    2024

  13. [21]

    Kevin Lin, Christopher Agia, Toki Migimatsu, Marco Pavone, and Jeannette Bohg. 2023. Text2motion: From natural language instructions to feasible plans.Autonomous Robots47, 8 (2023), 1345–1365

    2023

  14. [22]

    Jie Liu, Pan Zhou, Yingjun Du, Ah-Hwee Tan, Cees GM Snoek, Jan-Jakob Sonke, and Efstratios Gavves. 2024. CaPo: Cooperative Plan Optimization for Efficient Embodied Multi-Agent Cooperation.arXiv preprint arXiv:2411.04679(2024)

    2024 arXiv

  15. [23]

    Zhao Mandi, Shreeya Jain, and Shuran Song. 2024. Roco: Dialectic multi-robot collabora- tion with large language models. In2024 IEEE International Conference on Robotics and Automation. IEEE, 286–299

    2024

  16. [24]

    Siddharth Nayak, Adelmo Morrison Orozco, Marina Ten Have, Jackson Zhang, Vittal Thiru- malai, Darren Chen, Aditya Kapoor, Eric Robinson, Karthik Gopalakrishnan, James Harrison, et al. 2024. Long-Horizon Planning for Multi-Agent Robots in Partially Observable Environ- ments. In...

    2024

  17. [25]

    Kolby Nottingham, Prithviraj Ammanabrolu, Alane Suhr, Yejin Choi, Hannaneh Hajishirzi, Sameer Singh, and Roy Fox. 2023. Do embodied agents dream of pixelated sheep: Embodied de- cision making using language guided world modelling. InProceedings of the 40th International Confer...

    2023

  18. [26]

    2016.A concise introduction to decentralized POMDPs

    Frans A Oliehoek and Christopher Amato. 2016.A concise introduction to decentralized POMDPs. Springer

    2016

  19. [27]

    Xavier Puig, Tianmin Shu, Shuang Li, Zilin Wang, Yuan-Hong Liao, Joshua B Tenenbaum, Sanja Fidler, and Antonio Torralba. 2021. Watch-And-Help: A Challenge for Social Perception and Human-AI Collaboration. InProceedings of the 9th International Conference on Learning Representa...

    2021

  20. [28]

    Noah Shinn, Federico Cassano, Ashwin Gopinath, Karthik Narasimhan, and Shunyu Yao

  21. [29]

    David Silver and Richard S Sutton. 2025. Welcome to the era of experience

    2025

  22. [30]

    Chan Hee Song, Jiaman Wu, Clayton Washington, Brian M Sadler, Wei-Lun Chao, and Yu Su. 2023. Llm-planner: Few-shot grounded planning for embodied agents with large language models. InProceedings of the IEEE/CVF International Conference on Computer Vision. 2998– 3009

    2023

  23. [31]

    Jiankai Sun, Chuanyang Zheng, Enze Xie, Zhengying Liu, Ruihang Chu, Jianing Qiu, Jiaqi Xu, Mingyu Ding, Hongyang Li, Mengzhe Geng, et al . 2023. A Survey of Reasoning with Foundation Models: Concepts, Methodologies, and Outlook.Comput. Surveys(2023)

    2023

  24. [32]

    Gokul Swamy, Siddharth Reddy, Sergey Levine, and Anca D Dragan. 2020. Scaled autonomy: Enabling human operators to control robot fleets. In2020 IEEE International Conference on Robotics and Automation. IEEE, 5942–5948

    2020

  25. [33]

    Alejandro Torreno, Eva Onaindia, Antonín Komenda, and Michal Štolba. 2017. Cooperative multi-agent planning: A survey.ACM Computing Surveys (CSUR)50, 6 (2017), 1–32

    2017

  26. [34]

    2023.Llama: Open and efficient foundation language models

    Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timo- thée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. 2023.Llama: Open and efficient foundation language models. Technical Report. Meta AI

    2023

  27. [35]

    Zhenyu Wu, Ziwei Wang, Xiuwei Xu, Jiwen Lu, and Haibin Yan. 2023. Embodied task planning with large language models.arXiv preprint arXiv:2307.01848(2023)

    2023 arXiv

  28. [36]

    Zhuo Xu, Hao-Tien Lewis Chiang, Zipeng Fu, Mithun George Jacob, Tingnan Zhang, Tsang- Wei Edward Lee, Wenhao Yu, Connor Schenck, David Rendleman, Dhruv Shah, et al. 2024. Mobility VLA: Multimodal Instruction Navigation with Long-Context VLMs and Topological Graphs. InProceedin...

    2024

  29. [37]

    Ceyao Zhang, Kaijie Yang, Siyi Hu, Zihao Wang, Guanghe Li, Yihang Sun, Cheng Zhang, Zhaowei Zhang, Anji Liu, Song-Chun Zhu, et al. 2024. ProAgent: building proactive cooperative agents with large language models. InProceedings of the AAAI Conference on Artificial Intelligence,...

    2024

  30. [38]

    Hongxin Zhang, Weihua Du, Jiaming Shan, Qinhong Zhou, Yilun Du, Joshua B Tenenbaum, Tianmin Shu, and Chuang Gan. 2024. Building Cooperative Embodied Agents Modularly with Large Language Models. InProceedings of the 12th International Conference on Learning Representations

    2024

  31. [39]

    Gengze Zhou, Yicong Hong, Zun Wang, Xin Eric Wang, and Qi Wu. 2024. Navgpt-2: Unleash- ing navigational reasoning capability for large vision-language models. InProceedings of the 18th European Conference on Computer Vision. Springer, 260–278

    2024

  32. [40]

    Gengze Zhou, Yicong Hong, and Qi Wu. 2024. Navgpt: Explicit reasoning in vision-and- language navigation with large language models. InProceedings of the AAAI Conference on Artificial Intelligence, V ol. 38. 7641–7649

    2024

  33. [41]

    Hi, I’ll let you know if I find any goal objects and finish any subgoals, and ask for your help when necessary

    Ming Zhou, Jun Luo, Julian Villella, Yaodong Yang, David Rusu, Jiayu Miao, Weinan Zhang, Montgomery Alban, Iman Fadakar, Zheng Chen, et al. 2021. Smarts: An open-source scalable multi-agent rl training school for autonomous driving. InProceedings of the Conference on Robot Lea...

    2021

  34. [2023]

    InAdvances in Neural Information Processing Systems, V ol

    Reflexion: Language agents with verbal reinforcement learning. InAdvances in Neural Information Processing Systems, V ol. 36. 8634–8652

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.