The reviewed record of science sign in
Pith

arxiv: 2605.09423 · v2 · submitted 2026-05-10 · cs.AI

SimWorld Studio: Automatic Environment Generation with Evolving Coding Agent for Embodied Agent Learning

Haoqiang Kang , Xiaokang Ye , Yuhan Liu , Siddhant Hitesh Mantri , Lingjun Mao , James Fleming , Drishti Regmi , Lianhui Qin This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-05-14 21:26 UTCgrok-4.3open to challenge →

classification cs.AI
keywords embodied agentsenvironment generationcoding agentsself-evolutionco-evolution3D simulationnavigation tasksUnreal Engine
0
0 comments X

The pith

SimWorld Studio uses a self-evolving coding agent to generate adaptive 3D environments that raise embodied navigation success rates by 18 points over fixed training.

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

The paper introduces SimWorld Studio, a platform built on Unreal Engine 5 that automatically creates interactive 3D worlds for training embodied agents. At its core is SimCoder, which turns language or image instructions into engine-level code, then revises the output using feedback from compilation errors, physics checks, and visual critiques. The system also supports co-evolution, where the embodied agent's performance data directs the creation of progressively harder environments near the agent's current ability level. Three navigation case studies demonstrate that self-evolution makes environment generation more reliable, that the resulting worlds improve agent performance, and that these improvements transfer to unseen benchmarks. Co-evolution specifically delivers an 18-point success-rate increase compared with training in fixed environments and a 40-point increase relative to an untrained baseline.

Core claim

SimWorld Studio is an open-source platform on Unreal Engine 5 in which SimCoder, a tool-augmented coding agent, constructs physically grounded 3D environments from language or image instructions and exports them as Gym-style interfaces. SimCoder self-evolves by incorporating verifier feedback to fix errors and accumulate reusable tools and skills. The platform further enables co-evolution by feeding embodied-agent performance back to SimCoder so that it generates adaptive curricula at the learner's capability frontier.

What carries the argument

SimCoder, a tool- and skill-augmented coding agent that writes Unreal Engine 5 code and self-evolves by revising environments according to verifier feedback from compilation errors, physics checks, and VLM critiques.

If this is right

  • Self-evolution via verifier feedback measurably increases the reliability of generated 3D environments.
  • Environments produced by SimWorld Studio improve embodied-agent success rates on navigation tasks.
  • Performance gains from the generated environments transfer to unseen benchmarks.
  • Co-evolution between environment generation and agent learning produces an 18-point success-rate gain over fixed-environment training.
  • The same co-evolution process yields a 40-point success-rate gain relative to an untrained agent.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same verifier-driven loop could be ported to other game engines to expand the range of automatically generated training worlds.
  • Adaptive curricula near the agent's frontier may reduce the sample complexity of embodied learning compared with static datasets.
  • If the method scales, it offers a route to training regimes that require far less manual scene design than current embodied simulators.

Load-bearing premise

Feedback from compilation errors, physics checks, and visual-language critiques is sufficient to drive reliable self-evolution of complex, task-verifiable 3D environments without frequent human intervention.

What would settle it

A controlled run in which SimCoder receives repeated verifier feedback yet still produces non-functional or non-generalizing environments for a standard navigation benchmark, or in which co-evolution yields no measurable success-rate difference from fixed-environment training.

Figures

Figures reproduced from arXiv: 2605.09423 by Drishti Regmi, Haoqiang Kang, James Fleming, Lianhui Qin, Lingjun Mao, Siddhant Hitesh Mantri, Xiaokang Ye, Yuhan Liu.

Figure 1
Figure 1. Figure 1: SIMWORLD STUDIO: (Left) SIMCODER automatically generates UE5 interactive environments with realistic 3D scenes, learning tasks, and Gym interfaces. (Right) Co-evolving environment generation with embodied learning substantially improves test success over both fixed￾environment training and the untrained-agent baseline. Abstract LLM/VLM-based digital agents have advanced rapidly thanks to scalable sand￾boxe… view at source ↗
Figure 2
Figure 2. Figure 2: SIMCODER turns a user prompt into an interactive environment through an automatic self-evolving loop: it writes tools, creates reusable skills, reuses them across iterations, and refines the scene with verifier feedback. NavMesh-based tools are used to generate solvable navigation tasks. SIMCODER furthers uses embodied-agent feedback to autonomously adapt environment difficulty and co-evolve with the embod… view at source ↗
Figure 3
Figure 3. Figure 3: Three case studies evaluating SIMWORLD STUDIO. Case 1 evaluates SIMCODER’s scene generation quality across settings and LLM backbones. Case 2 trains embodied navigation agents in generated environments. Case 3 studies co-evolution where SIMCODER and the embodied agent iteratively improve each other. 3.1 Case Study 1: Can SIMCODER generate valid and diverse environments? This case study evaluates whether SI… view at source ↗
Figure 3
Figure 3. Figure 3: The workspace supports the full environment-generation workflow in a single view. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Ablation study results for SIMCODER in Case Study 1. Ablation studies on key platform components [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 4
Figure 4. Figure 4: Three case studies evaluating SIMWORLD STUDIO. Case 1 evaluates SIMCODER’s scene generation quality across settings and LLM backbones. Case 2 trains embodied navigation agents in generated environments. Case 3 studies co-evolution where SIMCODER and the embodied agent iteratively improve each other. 3.1 Case Study 1: Can SIMCODER generate valid and diverse environments? This case study evaluates whether SI… view at source ↗
Figure 5
Figure 5. Figure 5: Qualitative text-to-scene example. (Top) User prompt and rendered UE5 scenes from three model backbones. (a) The MCP tool spawn_blueprint_actor used throughout, showing its full interface: required parameters (actor_name, blueprint_id, location) and optional parameters (rotation, scale). (b) The Building Placement & Spacing skill retrieved by SIMCODER before generation; it provides building size categories… view at source ↗
Figure 5
Figure 5. Figure 5: Ablation study results for SIMCODER in Case Study 1. Ablation studies on key platform components [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Generalization analysis. (Left) More diverse SIMWORLD STUDIO environments yield stronger test-time generalization. (Right) Embodied agents learned in SIMWORLD STUDIO transfer to SimWorld-MMNav across model scales. 9 [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 6
Figure 6. Figure 6: Qualitative medieval village scene comparison. Each generated scene is paired with its corresponding text-to-scene evaluation scores. Scores are normalized to [0, 1]; higher is better. 8 [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Co-evolution of SIMCODER and embodied agent. (a) Environment difficulty across 8 levels. (b) Training dynamics: the co-evolving agent drops at each level transition then recovers. (c) Test performance on the SimWorld-MMNav benchmark. 3.3.1 Results Adaptive curricula drive continuous improvement and prevent early saturation. The training dynamics of the co-evolving system (Figure 7b) exhibit a characteristi… view at source ↗
Figure 7
Figure 7. Figure 7: Example of tool, skill, and generated script used in the example in [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Representative interface views of SIMWORLD STUDIO. The light-theme main interface provides an integrated workspace for user–agent interaction, UE scene rendering, asset/backend management, Gym environment APIs, and embodied-agent monitoring. The dark-theme panels further show specialized views for skill management, tool abstraction, and direct embodied interaction, allowing users to move beyond text-only p… view at source ↗
Figure 8
Figure 8. Figure 8: Generalization analysis. (Left) More diverse SIMWORLD STUDIO environments yield stronger test-time generalization. (Right) Embodied agents learned in SIMWORLD STUDIO transfer to SimWorld-MMNav across model scales. 10 [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Qualitative Example P1. Output scenes generated by three model backbones given the same downtown city-block intersection prompt. 40 [PITH_FULL_IMAGE:figures/full_fig_p040_9.png] view at source ↗
Figure 9
Figure 9. Figure 9: Co-evolution of SIMCODER and embodied agent. (a) Environment difficulty across 8 levels. (b) Training dynamics: the co-evolving agent drops at each level transition then recovers. (c) Test performance on the SimWorld-MMNav benchmark. 3.3.1 Results Adaptive curricula drive continuous improvement and prevent early saturation. The training dynamics of the co-evolving system (Figure 9b) exhibit a characteristi… view at source ↗
Figure 10
Figure 10. Figure 10: Qualitative Example P1. Top: prompt and reference image. Bottom: rendered UE5 screenshots from each model backbone. I.3 Scene Editing Prompt P1 Build a two-sided residential street in the current scene, which already has six starting buildings and six trees. Keep one existing building, remove the others, then fill out both sides using exactly two building types, the kept one plus one other medium sized bu… view at source ↗
Figure 10
Figure 10. Figure 10: Representative interface views of SIMWORLD STUDIO. The light-theme main interface provides an integrated workspace for user–agent interaction, UE scene rendering, asset/backend management, Gym environment APIs, and embodied-agent monitoring. The dark-theme panels further show specialized views for skill management, tool abstraction, and direct embodied interaction, allowing users to move beyond text-only … view at source ↗
Figure 11
Figure 11. Figure 11: Qualitative Example P1. Top: editing prompt and the original scene prior to modification. Bottom: rendered UE5 screenshots showing each model’s edited scene, built on top of the same starting configuration. 41 [PITH_FULL_IMAGE:figures/full_fig_p041_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Qualitative Example P2. Top: editing prompt and the original scene prior to modification. Bottom: rendered UE5 screenshots showing each model’s edited scene, built on top of the same starting configuration. I.4 Iterative Scene Development [PITH_FULL_IMAGE:figures/full_fig_p042_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Iterative scene development over six steps. Starting from a bare 4-way road intersection (Iter-1), the scene is progressively enriched through a sequence of natural language editing instructions: tall downtown buildings are added at each corner (Iter-2), sidewalks are dressed with trees and lamps (Iter-3), pedestrians populate the crosswalks and sidewalks (Iter-4), cars, scooters, and traffic signals are … view at source ↗
Figure 14
Figure 14. Figure 14: Qualitative Example P2. Top: editing prompt and the original scene prior to modification. Bottom: rendered UE5 screenshots showing each model’s edited scene, built on top of the same starting configuration. I.4 Iterative Scene Development [PITH_FULL_IMAGE:figures/full_fig_p043_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Iterative scene development over six steps. Starting from a bare 4-way road intersection (Iter-1), the scene is progressively enriched through a sequence of natural language editing instructions: tall downtown buildings are added at each corner (Iter-2), sidewalks are dressed with trees and lamps (Iter-3), pedestrians populate the crosswalks and sidewalks (Iter-4), cars, scooters, and traffic signals are … view at source ↗
read the original abstract

LLM/VLM-based digital agents have advanced rapidly thanks to scalable sandboxes for coding, web navigation, and computer use, which provide rich interactive training grounds. In contrast, embodied agents still lack abundant, diverse, and automatically generated 3D environments for interactive learning. Existing embodied simulators rely on manually crafted scenes or procedural templates, while recent LLM-based 3D generation systems mainly produce static scenes rather than deployable environments with verifiable tasks and standard learning interfaces. We introduce SimWorld Studio, an open-source platform built on Unreal Engine 5 for generating evolving embodied learning environments. At its core is SimCoder, a tool/skill-augmented coding agent that writes and executes engine-level code to construct physically grounded 3D worlds from language/image instructions. SimCoder self-evolves by using verifier feedback (e.g., compilation errors, physics checks, VLM critiques) to revise environments and autonomously add reusable tools and skills to its library. Generated worlds are exported as Gym-style environments for embodied agent learning. SimWorld Studio further enables co-evolution between environment generation and embodied learning: agent performance feedback guides SimCoder to generate adaptive curricula near the learner's capability frontier, so that environments become increasingly challenging as the embodied agent improves. Three case studies on embodied navigation show that self-evolution improves generation reliability, generated environments substantially improve embodied agent performance that generalizes to unseen benchmarks, and co-evolution yields an 18-point success-rate gain over fixed-environment learning and a 40-point gain over an untrained agent.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript introduces SimWorld Studio, an open-source platform on Unreal Engine 5 that employs a tool-augmented coding agent (SimCoder) to generate physically grounded 3D environments from language or image instructions. SimCoder self-evolves by incorporating verifier feedback such as compilation errors, physics checks, and VLM critiques, while also adding reusable tools to its library. Generated environments are exported as Gym-style interfaces. The platform further supports co-evolution, in which embodied agent performance feedback guides SimCoder to produce adaptive curricula near the learner's capability frontier. Three case studies on embodied navigation tasks report that self-evolution improves generation reliability, that the generated environments yield substantial performance gains with generalization to unseen benchmarks, and that co-evolution produces an 18-point success-rate improvement over fixed-environment learning and a 40-point improvement over an untrained agent.

Significance. If the reported gains hold, the work provides a concrete step toward scalable, automatically generated training environments for embodied agents, reducing dependence on manually crafted scenes. The open-source release, the use of standard Gym interfaces, and the demonstration of generalization to external benchmarks are positive features that could support reproducibility and follow-on research. The co-evolution loop offers a plausible mechanism for creating capability-matched curricula.

major comments (2)
  1. [Abstract and Case Studies] Abstract and Case Studies: The reported 18-point and 40-point success-rate gains are presented without accompanying information on the number of experimental runs, standard deviations or error bars, exact baseline implementations, or the precise protocol used to enforce task verifiability. These details are required to assess whether the generalization claims to unseen benchmarks are statistically robust.
  2. [Case Studies] Case Studies: No quantitative metrics are supplied on the autonomous evolution success rate, the number of self-revision iterations, or the frequency of human interventions needed to correct verifier failures. Because the central claim depends on verifier feedback (compilation errors, physics checks, VLM critiques) reliably producing task-verifiable environments with minimal manual correction, the absence of these statistics leaves the reliability of the self-evolution loop unsubstantiated.
minor comments (1)
  1. [Abstract] Abstract: On first use, the terms 'SimCoder' and 'SimWorld Studio' would benefit from a short parenthetical gloss to aid readers who encounter the abstract in isolation.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and positive overall assessment. We address each major comment below and have revised the manuscript to incorporate the requested details on experimental statistics and evolution metrics.

read point-by-point responses

  1. Referee: [Abstract and Case Studies] Abstract and Case Studies: The reported 18-point and 40-point success-rate gains are presented without accompanying information on the number of experimental runs, standard deviations or error bars, exact baseline implementations, or the precise protocol used to enforce task verifiability. These details are required to assess whether the generalization claims to unseen benchmarks are statistically robust.

    Authors: We agree that these statistical details are necessary to evaluate robustness. The revised manuscript expands the Case Studies section with the number of experimental runs, standard deviations and error bars, exact baseline implementations, and the full task verifiability protocol (combining compilation checks, physics validation, VLM critiques, and targeted human review). These additions confirm the statistical significance of the reported gains and support the generalization claims. revision: yes

  2. Referee: [Case Studies] Case Studies: No quantitative metrics are supplied on the autonomous evolution success rate, the number of self-revision iterations, or the frequency of human interventions needed to correct verifier failures. Because the central claim depends on verifier feedback (compilation errors, physics checks, VLM critiques) reliably producing task-verifiable environments with minimal manual correction, the absence of these statistics leaves the reliability of the self-evolution loop unsubstantiated.

    Authors: We acknowledge the importance of these metrics for substantiating the self-evolution claims. The revised manuscript adds a dedicated paragraph in the Case Studies section reporting quantitative metrics on the autonomous evolution success rate, average number of self-revision iterations, and frequency of human interventions. These statistics demonstrate that the verifier feedback loop produces reliable environments with limited manual correction. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical gains measured on external benchmarks and controls

full rationale

The paper describes an empirical system (SimCoder + Gym export + co-evolution loop) whose central claims are performance deltas on embodied navigation case studies. These deltas are reported against fixed-environment baselines, untrained agents, and unseen benchmarks; no equations, fitted parameters, or self-referential quantities are defined. No load-bearing self-citation, uniqueness theorem, or ansatz is invoked to derive the reported 18-point or 40-point gains. The derivation chain is therefore self-contained and externally falsifiable.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 2 invented entities

The system rests on the domain assumption that Unreal Engine 5 scripting can produce physically consistent, task-verifiable environments from generated code, plus the standard assumption that LLM-based agents can iteratively repair code using compilation and visual feedback. No explicit free parameters are introduced; the only invented entities are the platform and agent themselves, which are engineering artifacts rather than new physical postulates.

axioms (1)
  • domain assumption Unreal Engine 5 can be programmatically scripted to produce physically grounded, interactive 3D scenes with verifiable tasks
    Invoked as the foundation for SimCoder's code execution and export to Gym environments.
invented entities (2)
  • SimCoder no independent evidence
    purpose: Tool-augmented coding agent that writes, executes, and self-improves environment-generation code
    Core new component introduced to automate environment creation; no independent falsifiable prediction outside the system is provided.
  • SimWorld Studio no independent evidence
    purpose: Integrated platform enabling self-evolution and co-evolution of environments with embodied learners
    The overall system and its co-evolution loop are new engineering contributions.

pith-pipeline@v0.9.0 · 5605 in / 1584 out tokens · 45812 ms · 2026-05-14T21:26:27.808425+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

126 extracted references · 126 canonical work pages · 30 internal anchors

  1. [1]

    GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning

    Lakshya A Agrawal, Shangyin Tan, Dilara Soylu, Noah Ziems, Rishi Khare, Krista Opsahl- Ong, Arnav Singhvi, Herumb Shandilya, Michael J Ryan, Meng Jiang, Christopher Potts, 12 Koushik Sen, Alexandros G. Dimakis, Ion Stoica, Dan Klein, Matei Zaharia, and Omar Khattab. Gepa: Reflective prompt evolution can outperform reinforcement learning, 2026. URLhttps://...

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  2. [2]

    Peter Anderson, Angel Chang, Devendra Singh Chaplot, Alexey Dosovitskiy, Saurabh Gupta, Vladlen Koltun, Jana Kosecka, Jitendra Malik, Roozbeh Mottaghi, Manolis Savva, and Amir R. Zamir. On evaluation of embodied navigation agents, 2018. URL https://arxiv.org/ abs/1807.06757

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  3. [3]

    Introducing the model context protocol.https://www.anthropic.com/news/ model-context-protocol, 2024

    Anthropic. Introducing the model context protocol.https://www.anthropic.com/news/ model-context-protocol, 2024. Published November 25, 2024. Accessed: 2026-05-03

    work page 2024

  4. [4]

    Claude code by anthropic

    Anthropic. Claude code by anthropic. https://www.anthropic.com/product/ claude-code, 2026. Accessed: 2026-05-07

    work page 2026

  5. [5]

    Introducing claude opus 4.6

    Anthropic. Introducing claude opus 4.6. https://www.anthropic.com/news/ claude-opus-4-6, February 2026. Accessed: 2026-05-07

    work page 2026

  6. [6]

    Claude sonnet 4.6

    Anthropic. Claude sonnet 4.6. https://www.anthropic.com/claude/sonnet, February

    work page

  7. [7]

    Accessed: 2026-05-07

    work page 2026

  8. [8]

    Program Synthesis with Large Language Models

    Jacob Austin, Augustus Odena, Maxwell Nye, Maarten Bosma, Henryk Michalewski, David Dohan, Ellen Jiang, Carrie Cai, Michael Terry, Quoc Le, et al. Program synthesis with large language models.arXiv preprint arXiv:2108.07732, 2021. URL https://arxiv.org/abs/ 2108.07732

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  9. [9]

    Objectnav revisited: On evaluation of embodied agents navigating to objects

    Dhruv Batra, Aaron Gokaslan, Aniruddha Kembhavi, Oleksandr Maksymets, Roozbeh Mot- taghi, Manolis Savva, Alexander Toshev, and Erik Wijmans. Objectnav revisited: On evaluation of embodied agents navigating to objects.CoRR, abs/2006.13171, 2020. URL https://arxiv.org/abs/2006.13171

    work page arXiv 2006

  10. [10]

    Scenefactor: Factored latent 3d diffusion for controllable 3d scene generation

    Aleksey Bokhovkin, Quan Meng, Shubham Tulsiani, and Angela Dai. Scenefactor: Factored latent 3d diffusion for controllable 3d scene generation. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 628–639, 2025

    work page 2025

  11. [11]

    Genie: Generative interactive environments

    Jake Bruce, Michael D Dennis, Ashley Edwards, Jack Parker-Holder, Yuge Shi, Edward Hughes, Matthew Lai, Aditi Mavalankar, Richie Steigerwald, Chris Apps, et al. Genie: Generative interactive environments. InForty-first International Conference on Machine Learning, 2024

    work page 2024

  12. [12]

    Large language models as tool makers, 2024

    Tianle Cai, Xuezhi Wang, Tengyu Ma, Xinyun Chen, and Denny Zhou. Large language models as tool makers, 2024

    work page 2024

  13. [13]

    Evaluating Large Language Models Trained on Code

    Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Ponde De Oliveira Pinto, Jared Kaplan, Harri Edwards, Yuri Burda, Nicholas Joseph, Greg Brockman, et al. Evaluating large language models trained on code.arXiv preprint arXiv:2107.03374, 2021. URL https://arxiv.org/abs/2107.03374

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  14. [14]

    Creator: Tool creation for disentangling abstract and concrete reasoning of large language models, 2023

    Qian Cheng, Xin Cong, Zhong Zhang, Yesai Wu, Yankai Lin, Xie Zhiyuan, Liu Zhiyuan, and Sun Maosong. Creator: Tool creation for disentangling abstract and concrete reasoning of large language models, 2023

    work page 2023

  15. [15]

    LEGENT: open platform for embodied agents,

    Zhili Cheng, Zhitong Wang, Jinyi Hu, Shengding Hu, An Liu, Yuge Tu, Pengkai Li, Lei Shi, Zhiyuan Liu, and Maosong Sun. LEGENT: Open platform for embodied agents. In Yixin Cao, Yang Feng, and Deyi Xiong, editors,Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), pages 335–345, Bangkok,...

    work page doi:10.18653/v1/2024.acl-demos.32 2024

  16. [16]

    arXiv preprint arXiv:2311.13384 , year=

    Jaeyoung Chung, Suyoung Lee, Hyeongjin Nam, Jaerin Lee, and Kyoung Mu Lee. Lu- ciddreamer: Domain-free generation of 3d gaussian splatting scenes, 2023. URL https: //arxiv.org/abs/2311.13384. 13

    work page arXiv 2023

  17. [17]

    Procthor: Large-scale embodied ai using procedural generation.arXiv preprint arXiv:2206.06994, 2022

    Matt Deitke, Eli VanderBilt, Alvaro Herrasti, Luca Weihs, Jordi Salvador, Kiana Ehsani, Winson Han, Eric Kolve, Ali Farhadi, Aniruddha Kembhavi, and Roozbeh Mottaghi. Procthor: Large-scale embodied ai using procedural generation, 2022. URL https://arxiv.org/ abs/2206.06994

    work page arXiv 2022

  18. [18]

    Emergent complexity and zero-shot transfer via unsupervised environment design.Advances in neural information processing systems, 33:13049–13061, 2020

    Michael Dennis, Natasha Jaques, Eugene Vinitsky, Alexandre Bayen, Stuart Russell, Andrew Critch, and Sergey Levine. Emergent complexity and zero-shot transfer via unsupervised environment design.Advances in neural information processing systems, 33:13049–13061, 2020

    work page 2020

  19. [19]

    Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence

    Guanting Dong, Junting Lu, Junjie Huang, Wanjun Zhong, Longxiang Liu, Shijue Huang, Zhenyu Li, Yang Zhao, Xiaoshuai Song, Xiaoxi Li, Jiajie Jin, Yutao Zhu, Hanbin Wang, Fangyu Lei, Qinyu Luo, Mingyang Chen, Zehui Chen, Jiazhan Feng, Ji-Rong Wen, and Zhicheng Dou. Agent-world: Scaling real-world environment synthesis for evolving general agent intelligence...

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  20. [20]

    CARLA: An Open Urban Driving Simulator

    Alexey Dosovitskiy, German Ros, Felipe Codevilla, Antonio Lopez, and Vladlen Koltun. Carla: An open urban driving simulator, 2017. URLhttps://arxiv.org/abs/1711.03938

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  21. [21]

    Danny Driess, Fei Xia, Mehdi S. M. Sajjadi, Corey Lynch, Aakanksha Chowdhery, Brian Ichter, Ayzaan Wahid, Jonathan Tompson, Quan Vuong, Tianhe Yu, Wenlong Huang, Yevgen Chebotar, Pierre Sermanet, Daniel Duckworth, Sergey Levine, Vincent Vanhoucke, Karol Hausman, Marc Toussaint, Klaus Greff, Andy Zeng, Igor Mordatch, and Pete Florence. Palm-e: An embodied ...

    work page 2023

  22. [22]

    Omni-epic: Open- endedness via models of human notions of interestingness with environ- ments programmed in code,

    Maxence Faldor, Jenny Zhang, Antoine Cully, and Jeff Clune. Omni-epic: Open-endedness via models of human notions of interestingness with environments programmed in code.arXiv preprint arXiv:2405.15568, 2024. URLhttps://arxiv.org/abs/2405.15568

    work page arXiv 2024

  23. [23]

    Minedojo: Building open-ended embodied agents with internet-scale knowledge.arXiv preprint arXiv:2206.08853, 2022

    Linxi Fan, Guanzhi Wang, Yunfan Jiang, Ajay Mandlekar, Yuncong Yang, Haoyi Zhu, Andrew Tang, De-An Huang, Yuke Zhu, and Anima Anandkumar. Minedojo: Building open-ended embodied agents with internet-scale knowledge, 2022. URL https://arxiv.org/abs/ 2206.08853

    work page arXiv 2022

  24. [24]

    A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems

    Jinyuan Fang, Yanwen Peng, Xi Zhang, Yingxu Wang, Xinhao Yi, Guibin Zhang, Yi Xu, Bin Wu, Siwei Liu, Zihao Li, et al. A comprehensive survey of self-evolving ai agents: A new paradigm bridging foundation models and lifelong agentic systems.arXiv preprint arXiv:2508.07407, 2025. URLhttps://arxiv.org/abs/2508.07407

    work page internal anchor Pith review arXiv 2025

  25. [25]

    Webevolver: Enhancing web agent self-improvement with co-evolving world model

    Tianqing Fang, Hongming Zhang, Zhisong Zhang, Kaixin Ma, Wenhao Yu, Haitao Mi, and Dong Yu. Webevolver: Enhancing web agent self-improvement with co-evolving world model. InProceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 8970–8986, 2025

    work page 2025

  26. [26]

    ThreeDWorld: A Platform for Interactive Multi-Modal Physical Simulation

    Chuang Gan, Jeremy Schwartz, Seth Alter, Damian Mrowca, Martin Schrimpf, James Traer, Julian De Freitas, Jonas Kubilius, Abhishek Bhandwaldar, Nick Haber, Megumi Sano, Kuno Kim, Elias Wang, Michael Lingelbach, Aidan Curtis, Kevin Feigelis, Daniel M. Bear, Dan Gutfreund, David Cox, Antonio Torralba, James J. DiCarlo, Joshua B. Tenenbaum, Josh H. McDermott,...

    work page Pith review arXiv 2021

  27. [27]

    Embodiedcity: A benchmark platform for embodied agent in real-world city environment.arXiv preprint, 2024

    Chen Gao, Baining Zhao, Weichen Zhang, Jun Zhang, Jinzhu Mao, Zhiheng Zheng, Fanhang Man, Jianjie Fang, Zile Zhou, Jinqiang Cui, Xinlei Chen, and Yong Li. Embodiedcity: A benchmark platform for embodied agent in real-world city environment.arXiv preprint, 2024

    work page 2024

  28. [28]

    A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence

    Huan-ang Gao, Jiayi Geng, Wenyue Hua, Mengkang Hu, Xinzhe Juan, Hongzhang Liu, Shilong Liu, Jiahao Qiu, Xuan Qi, Yiran Wu, et al. A survey of self-evolving agents: What, when, how, and where to evolve on the path to artificial super intelligence.arXiv preprint arXiv:2507.21046, 2025. URLhttps://arxiv.org/abs/2507.21046. 14

    work page internal anchor Pith review arXiv 2025

  29. [29]

    Genenv: Difficulty-aligned co-evolution between llm agents and environment simulators.arXiv preprint arXiv:2512.19682, 2025a

    Jiacheng Guo, Ling Yang, Peter Chen, Qixin Xiao, Yinjie Wang, Xinzhe Juan, Jiahao Qiu, Ke Shen, and Mengdi Wang. Genenv: Difficulty-aligned co-evolution between llm agents and environment simulators.arXiv preprint arXiv:2512.19682, 2025. URL https://arxiv. org/abs/2512.19682

    work page arXiv 2025

  30. [30]

    Vlaw: Iterative co-improvement of vision-language-action policy and world model.CoRR, abs/2602.12063, 2026

    Yanjiang Guo, Tony Lee, Lucy Xiaoyang Shi, Jianyu Chen, Percy Liang, and Chelsea Finn. Vlaw: Iterative co-improvement of vision-language-action policy and world model.arXiv preprint arXiv:2602.12063, 2026. URLhttps://arxiv.org/abs/2602.12063

    work page arXiv 2026

  31. [31]

    Openwebvoyager: Building multimodal web agents via iterative real-world exploration, feedback and optimization

    Hongliang He, Wenlin Yao, Kaixin Ma, Wenhao Yu, Hongming Zhang, Tianqing Fang, Zhenzhong Lan, and Dong Yu. Openwebvoyager: Building multimodal web agents via iterative real-world exploration, feedback and optimization. InProceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 27545–27564, 2025

    work page 2025

  32. [32]

    Text2room: Extracting textured 3d meshes from 2d text-to-image models

    Lukas Höllein, Ang Cao, Andrew Owens, Justin Johnson, and Matthias Nießner. Text2room: Extracting textured 3d meshes from 2d text-to-image models. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 7909–7920, 2023

    work page 2023

  33. [33]

    Metagpt: Meta programming for a multi-agent collaborative framework

    Sirui Hong, Mingchen Zhuge, Jonathan Chen, Xiawu Zheng, Yuheng Cheng, Jinlin Wang, Ceyao Zhang, Zili Wang, Steven Ka Shing Yau, Zijuan Lin, et al. Metagpt: Meta programming for a multi-agent collaborative framework. InThe twelfth international conference on learning representations, 2023

    work page 2023

  34. [34]

    Relic: Interactive video world model with long-horizon memory.arXiv preprint arXiv:2512.04040, 2025

    Yicong Hong, Yiqun Mei, Chongjian Ge, Yiran Xu, Yang Zhou, Sai Bi, Yannick Hold- Geoffroy, Mike Roberts, Matthew Fisher, Eli Shechtman, et al. Relic: Interactive video world model with long-horizon memory.arXiv preprint arXiv:2512.04040, 2025. URL https://arxiv.org/abs/2512.04040

    work page arXiv 2025

  35. [35]

    Automated Design of Agentic Systems

    Shengran Hu, Cong Lu, and Jeff Clune. Automated design of agentic systems.arXiv preprint arXiv:2408.08435, 2024. URLhttps://arxiv.org/abs/2408.08435

    work page internal anchor Pith review arXiv 2024

  36. [36]

    Ross, Cordelia Schmid, and Alireza Fathi

    Ziniu Hu, Ahmet Iscen, Aashi Jain, Thomas Kipf, Yisong Yue, David A. Ross, Cordelia Schmid, and Alireza Fathi. Scenecraft: An llm agent for synthesizing 3d scene as blender code, 2024. URLhttps://arxiv.org/abs/2403.01248

    work page arXiv 2024

  37. [37]

    R-zero: Self-evolving reasoning llm from zero data

    Chengsong Huang, Wenhao Yu, Xiaoyang Wang, Hongming Zhang, Zongxia Li, Ruosen Li, Jiaxin Huang, Haitao Mi, and Dong Yu. R-zero: Self-evolving reasoning llm from zero data

    work page

  38. [38]

    URLhttps://arxiv.org/abs/2508.05004

    work page internal anchor Pith review Pith/arXiv arXiv

  39. [39]

    Prioritized Level Replay, 2020

    Minqi Jiang, Edward Grefenstette, and Tim Rocktäschel. Prioritized Level Replay, 2020

    work page 2020

  40. [40]

    WoVR: World Models as Reliable Simulators for Post-Training VLA Policies with RL

    Zhennan Jiang, Shangqing Zhou, Yutong Jiang, Zefang Huang, Mingjie Wei, Yuhui Chen, Tianxing Zhou, Zhen Guo, Hao Lin, Quanlu Zhang, Yu Wang, Haoran Li, Chao Yu, and Dongbin Zhao. Wovr: World models as reliable simulators for post-training vla policies with rl.arXiv preprint arXiv:2602.13977, 2026. URLhttps://arxiv.org/abs/2602.13977

    work page internal anchor Pith review arXiv 2026

  41. [41]

    SWE-bench: Can Language Models Resolve Real-World GitHub Issues?

    Carlos E Jimenez, John Yang, Alexander Wettig, Shunyu Yao, Kexin Pei, Ofir Press, and Karthik Narasimhan. Swe-bench: Can language models resolve real-world github issues? arXiv preprint arXiv:2310.06770, 2023. URLhttps://arxiv.org/abs/2310.06770

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  42. [42]

    AI2-THOR: An Interactive 3D Environment for Visual AI

    Eric Kolve, Roozbeh Mottaghi, Winson Han, Eli VanderBilt, Luca Weihs, Alvaro Herrasti, Matt Deitke, Kiana Ehsani, Daniel Gordon, Yuke Zhu, Aniruddha Kembhavi, Abhinav Gupta, and Ali Farhadi. Ai2-thor: An interactive 3d environment for visual ai, 2022. URL https: //arxiv.org/abs/1712.05474

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  43. [43]

    Vulcan: Tool-augmented multi agents for iterative 3d object arrangement, 2026

    Zhengfei Kuang, Rui Lin, Long Zhao, Gordon Wetzstein, Saining Xie, and Sanghyun Woo. Vulcan: Tool-augmented multi agents for iterative 3d object arrangement, 2026. URL https: //arxiv.org/abs/2512.22351

    work page arXiv 2026

  44. [45]

    URLhttps://arxiv.org/abs/2108.03272. 15

    work page arXiv

  45. [46]

    BEHAVIOR-1K: A Human-Centered, Embodied AI Benchmark with 1,000 Everyday Activities and Realistic Simulation

    Chengshu Li, Ruohan Zhang, Josiah Wong, Cem Gokmen, Sanjana Srivastava, Roberto Martín- Martín, Chen Wang, Gabrael Levine, Wensi Ai, Benjamin Martinez, Hang Yin, Michael Lingelbach, Minjune Hwang, Ayano Hiranaka, Sujay Garlanka, Arman Aydin, Sharon Lee, Jiankai Sun, Mona Anvari, Manasi Sharma, Dhruva Bansal, Samuel Hunter, Kyu-Young Kim, Alan Lou, Caleb R...

    work page internal anchor Pith review arXiv 2024

  46. [47]

    Eurekaverse: Environment Curriculum Generation via Large Language Models , journal =

    William Liang, Sam Wang, Hung-Ju Wang, Osbert Bastani, Dinesh Jayaraman, and Yecheng Ja- son Ma. Eurekaverse: Environment curriculum generation via large language models.arXiv preprint arXiv:2411.01775, 2024. URLhttps://arxiv.org/abs/2411.01775

    work page arXiv 2024

  47. [48]

    AgentBench: Evaluating LLMs as Agents

    Xiao Liu, Hao Yu, Hanchen Zhang, Yifan Xu, Xuanyu Lei, Hanyu Lai, Yu Gu, Hangliang Ding, Kaiwen Men, Kejuan Yang, et al. Agentbench: Evaluating llms as agents.arXiv preprint arXiv:2308.03688, 2023. URLhttps://arxiv.org/abs/2308.03688

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  48. [50]

    URLhttps://arxiv.org/abs/2602.06508

    work page internal anchor Pith review arXiv

  49. [51]

    Imagine a city: Citygenagent for procedural 3d city generation,

    Zishan Liu, Zecong Tang, RuoCheng Wu, Xinzhe Zheng, Jingyu Hu, Ka-Hei Hui, Haoran Xie, Bo Dai, and Zhengzhe Liu. Imagine a city: Citygenagent for procedural 3d city generation,

    work page

  50. [52]

    URLhttps://arxiv.org/abs/2602.05362

    work page arXiv

  51. [53]

    Self-adapting improvement loops for robotic learning.arXiv preprint arXiv:2506.06658, 2025

    Calvin Luo, Zilai Zeng, Mingxi Jia, Yilun Du, and Chen Sun. Self-adapting improvement loops for robotic learning.arXiv preprint arXiv:2506.06658, 2025. URL https://arxiv. org/abs/2506.06658

    work page arXiv 2025

  52. [54]

    Eureka: Human-Level Reward Design via Coding Large Language Models

    Yecheng Jason Ma, William Liang, Guanzhi Wang, De-An Huang, Osbert Bastani, Dinesh Jayaraman, Yuke Zhu, Linxi Fan, and Anima Anandkumar. Eureka: Human-level reward design via coding large language models.arXiv preprint arXiv:2310.12931, 2023. URL https://arxiv.org/abs/2310.12931

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  53. [55]

    Self-refine: Iterative refinement with self-feedback.Advances in neural information processing systems, 36:46534–46594, 2023

    Aman Madaan, Niket Tandon, Prakhar Gupta, Skyler Hallinan, Luyu Gao, Sarah Wiegreffe, Uri Alon, Nouha Dziri, Shrimai Prabhumoye, Yiming Yang, et al. Self-refine: Iterative refinement with self-feedback.Advances in neural information processing systems, 36:46534–46594, 2023

    work page 2023

  54. [56]

    Cosmos World Foundation Model Platform for Physical AI

    NVIDIA et al. Cosmos world foundation model platform for physical ai.arXiv preprint arXiv:2501.03575, 2025. URLhttps://arxiv.org/abs/2501.03575

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  55. [57]

    Generative Agents: Interactive Simulacra of Human Behavior

    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, 2023. URL https://arxiv.org/abs/2304.03442

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  56. [58]

    Evolving curricula with regret-based environment design

    Jack Parker-Holder, Minqi Jiang, Michael Dennis, Mikayel Samvelyan, Jakob Foerster, Edward Grefenstette, and Tim Rocktäschel. Evolving curricula with regret-based environment design. InInternational Conference on Machine Learning, pages 17473–17498. PMLR, 2022

    work page 2022

  57. [59]

    Atiss: Autoregressive transformers for indoor scene synthesis.Advances in neural information processing systems, 34:12013–12026, 2021

    Despoina Paschalidou, Amlan Kar, Maria Shugrina, Karsten Kreis, Andreas Geiger, and Sanja Fidler. Atiss: Autoregressive transformers for indoor scene synthesis.Advances in neural information processing systems, 34:12013–12026, 2021

    work page 2021

  58. [60]

    SceneSmith: Agentic Generation of Simulation-Ready Indoor Scenes

    Nicholas Pfaff, Thomas Cohn, Sergey Zakharov, Rick Cory, and Russ Tedrake. Scenesmith: Agentic generation of simulation-ready indoor scenes.arXiv preprint arXiv:2602.09153, 2026. URLhttps://arxiv.org/abs/2602.09153

    work page internal anchor Pith review arXiv 2026

  59. [61]

    Habitat 3.0: A Co-Habitat for Humans, Avatars and Robots

    Xavier Puig, Eric Undersander, Andrew Szot, Mikael Dallaire Cote, Tsung-Yen Yang, Ruslan Partsey, Ruta Desai, Alexander William Clegg, Michal Hlavac, So Yeon Min, Vladimír V ondruš, Theophile Gervet, Vincent-Pierre Berges, John M. Turner, Oleksandr Maksymets, Zsolt Kira, Mrinal Kalakrishnan, Jitendra Malik, Devendra Singh Chaplot, Unnat Jain, Dhruv Batra,...

    work page Pith review arXiv 2023

  60. [62]

    arXiv preprint arXiv:2411.02337 , year=

    Zehan Qi, Xiao Liu, Iat Long Iong, Hanyu Lai, Xueqiao Sun, Wenyi Zhao, Yu Yang, Xinyue Yang, Jiadai Sun, Shuntian Yao, et al. Webrl: Training llm web agents via self-evolving online curriculum reinforcement learning.arXiv preprint arXiv:2411.02337, 2024. URL https://arxiv.org/abs/2411.02337

    work page arXiv 2024

  61. [63]

    Toolllm: Facilitating large language models to master 16000+ real-world apis, 2023

    Yujia Qin, Shihao Liang, Yining Ye, Kunlun Zhu, Lan Yan, Yaxi Lu, Yankai Lin, Xin Cong, Xiangru Tang, Bill Qian, Sihan Zhao, Lauren Hong, Runchu Tian, Ruobing Xie, Jie Zhou, Mark Gerstein, Dahai Li, Zhiyuan Liu, and Maosong Sun. Toolllm: Facilitating large language models to master 16000+ real-world apis, 2023. URL https://arxiv.org/abs/2307. 16789

    work page 2023

  62. [64]

    Qwen3.5: Towards native multimodal agents, February 2026

    Qwen Team. Qwen3.5: Towards native multimodal agents, February 2026. URL https: //qwen.ai/blog?id=qwen3.5

    work page 2026

  63. [65]

    Infinite photorealistic worlds using procedural generation

    Alexander Raistrick, Lahav Lipson, Zeyu Ma, Lingjie Mei, Mingzhe Wang, Yiming Zuo, Karhan Kayan, Hongyu Wen, Beining Han, Yihan Wang, et al. Infinite photorealistic worlds using procedural generation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12630–12641, 2023

    work page 2023

  64. [66]

    A self-improving coding agent

    Maxime Robeyns, Martin Szummer, and Laurence Aitchison. A self-improving coding agent. arXiv preprint arXiv:2504.15228, 2025. URLhttps://arxiv.org/abs/2504.15228

    work page arXiv 2025

  65. [67]

    2023 , month = mar, number =

    Mikayel Samvelyan, Akbir Khan, Michael Dennis, Minqi Jiang, Jack Parker-Holder, Jakob Foerster, Roberta Raileanu, and Tim Rocktäschel. Maestro: Open-ended environment design for multi-agent reinforcement learning.arXiv preprint arXiv:2303.03376, 2023. URL https: //arxiv.org/abs/2303.03376

    work page arXiv 2023

  66. [68]

    Toolformer: Language Models Can Teach Themselves to Use Tools

    Timo Schick, Jane Dwivedi-Yu, Roberto Dessì, Roberta Raileanu, Maria Lomeli, Luke Zettle- moyer, Nicola Cancedda, and Thomas Scialom. Toolformer: Language models can teach themselves to use tools, 2023. URLhttps://arxiv.org/abs/2302.04761

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  67. [69]

    Proximal Policy Optimization Algorithms

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. Proximal policy optimization algorithms, 2017. URLhttps://arxiv.org/abs/1707.06347

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  68. [70]

    World-gymnast: Training robots with reinforcement learning in a world model.CoRR, abs/2602.02454, 2026

    Ansh Kumar Sharma, Yixiang Sun, Ninghao Lu, Yunzhe Zhang, Jiarao Liu, and Sherry Yang. World-gymnast: Training robots with reinforcement learning in a world model, 2026. URL https://arxiv.org/abs/2602.02454

    work page arXiv 2026

  69. [71]

    Re- flexion: Language agents with verbal reinforcement learning.Advances in neural information processing systems, 36:8634–8652, 2023

    Noah Shinn, Federico Cassano, Ashwin Gopinath, Karthik Narasimhan, and Shunyu Yao. Re- flexion: Language agents with verbal reinforcement learning.Advances in neural information processing systems, 36:8634–8652, 2023

    work page 2023

  70. [72]

    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, 2021. URLhttps://arxiv.org/abs/2010.03768

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  71. [73]

    Virtualenv: A platform for embodied ai research, 2026

    Kabir Swain, Sijie Han, Ayush Raina, Jin Zhang, Shuang Li, Michael Stopa, and Antonio Torralba. Virtualenv: A platform for embodied ai research, 2026. URL https://arxiv. org/abs/2601.07553

    work page arXiv 2026

  72. [74]

    Diffuscene: Denoising diffusion models for generative indoor scene synthesis

    Jiapeng Tang, Yinyu Nie, Lev Markhasin, Angela Dai, Justus Thies, and Matthias Nießner. Diffuscene: Denoising diffusion models for generative indoor scene synthesis. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 20507–20518, 2024

    work page 2024

  73. [75]

    Gigabrain-0.5 m*: a vla that learns from world model-based reinforcement learning.arXiv preprint arXiv:2602.12099, 2026

    GigaBrain Team, Boyuan Wang, Bohan Li, Chaojun Ni, et al. Gigabrain-0.5m*: a vla that learns from world model-based reinforcement learning.arXiv preprint arXiv:2602.12099,

    work page arXiv

  74. [76]

    Gigabrain-0.5 m*: a vla that learns from world model-based reinforcement learning.arXiv preprint arXiv:2602.12099, 2026

    doi: 10.48550/arXiv.2602.12099. URLhttps://arxiv.org/abs/2602.12099

    work page doi:10.48550/arxiv.2602.12099

  75. [77]

    Vygotsky.Mind in Society: The Development of Higher Psychological Processes

    Lev S. Vygotsky.Mind in Society: The Development of Higher Psychological Processes. Harvard University Press, Cambridge, MA, 1978. 17

    work page 1978

  76. [78]

    V oyager: An open-ended embodied agent with large language models, 2023

    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, 2023

    work page 2023

  77. [79]

    arXiv preprint arXiv:2407.10943 , year =

    Hanqing Wang, Jiahe Chen, Wensi Huang, Qingwei Ben, Tai Wang, Boyu Mi, Tao Huang, Siheng Zhao, Yilun Chen, Sizhe Yang, Peizhou Cao, Wenye Yu, Zichao Ye, Jialun Li, Junfeng Long, Zirui Wang, Huiling Wang, Ying Zhao, Zhongying Tu, Yu Qiao, Dahua Lin, and Jiangmiao Pang. Grutopia: Dream general robots in a city at scale, 2024. URL https: //arxiv.org/abs/2407.10943

    work page arXiv 2024

  78. [80]

    Rui Wang, Joel Lehman, Jeff Clune, and Kenneth O. Stanley. POET: Open-Ended Co- evolution of Environments and Their Optimized Solutions. InProceedings of the Genetic and Evolutionary Computation Conference (GECCO), pages 142–151. ACM, 2019. doi: 10.1145/3321707.3321799

    work page doi:10.1145/3321707.3321799 2019

  79. [81]

    Enhanced poet: Open-ended reinforcement learning through unbounded invention of learning challenges and their solutions

    Rui Wang, Joel Lehman, Aditya Rawal, Jiale Zhi, Yulun Li, Jeff Clune, and Kenneth O Stanley. Enhanced poet: Open-ended reinforcement learning through unbounded invention of learning challenges and their solutions. InProceedings of the 37th International Conference on Machine Learning, pages 9940–9951. PMLR, 2020. URL http://proceedings.mlr. press/v119/wan...

    work page 2020

  80. [82]

    Executable code actions elicit better llm agents

    Xingyao Wang, Yangyi Chen, Lifan Yuan, Yizhe Zhang, Yunzhu Li, Hao Peng, and Heng Ji. Executable code actions elicit better llm agents. InForty-first International Conference on Machine Learning, 2024

    work page 2024

Showing first 80 references.