Debate2Create: Robot Co-design via Multi-Agent LLM Debate
On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks
Abstract (arXiv)
We introduce Debate2Create (D2C), a multi-agent LLM framework that formulates robot co-design as structured, iterative debate grounded in physics-based evaluation. A design agent and control agent engage in a thesis-antithesis-synthesis loop, while criterion-specific LLM judges provide multi-objective feedback to steer exploration. Across five MuJoCo locomotion benchmarks, D2C achieves the highest default-normalized score among the evaluated LLM-based and black-box baselines, with gains up to 3.2x on Ant and nearly 9x on Swimmer. Iterative debate yields 18-35% gains over compute-matched zero-shot generation, and D2C-generated rewards transfer to default morphologies in 4/5 tasks. These results suggest that structured, simulator-grounded multi-agent interaction is a useful mechanism for joint morphology-reward optimization under a fixed-topology, per-candidate-RL protocol. Project page: debate2create.github.io.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 25 of its 52 references that are in the map. First posted on arXiv 29 Oct 2025.
Builds on
25 of 52 references.
Closest ancestors · 1
Same direction, newest first
Other references · 24
- 2017Proximal Policy Optimization Algorithmscited by 14,441
- 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,635
- 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,064
- 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,317
- 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,739
- 2016OpenAI Gymcited by 1,684
- 2016Concrete Problems in AI Safetycited by 1,582
- 2024A Survey on LLM-as-a-Judgecited by 1,384
- 2023Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debatecited by 995
- 2023Teaching Large Language Models to Self-Debugcited by 963
- 2023Eureka: Human-Level Reward Design via Coding Large Language Modelscited by 542
- 2015Illuminating search spaces by mapping elitescited by 406
12 more
- 2018AI safety via debatecited by 384
- 2021Brax -- A Differentiable Physics Engine for Large Scale Rigid Body Simulationcited by 375
- 2023Language to Rewards for Robotic Skill Synthesiscited by 307
- 2023Reward Design with Language Modelscited by 257
- 2023Text2Reward: Reward Shaping with Language Models for Reinforcement Learningcited by 153
- 2024DrEureka: Language Model Guided Sim-To-Real Transfercited by 78
- 2024The Perfect Blend: Redefining RLHF with Mixture of Judgescited by 43
- 2017Scalable Co-Optimization of Morphology and Control in Embodied Machinescited by 38
- 2025VLMgineer: Vision Language Models as Robotic Toolsmithscited by 9
- 2024Text2Robot: Evolutionary Robot Design from Text Descriptionscited by 8
- 2024RoboMorph: Evolving Robot Morphology using Large Language Modelscited by 8
- 2025RoboMoRe: LLM-based Robot Co-design via Joint Optimization of Morphology and Rewardcited by 4
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2022C^2:Co-design of Robots via Concurrent Networks Coupling Online and Offline Reinforcement Learning
- 2026LLawCo: Learning Laws of Cooperation for Modeling Embodied Multi-Agent Behavior
- 2026When Search Becomes Memory: Accelerating Robot Design Discovery with Self-Evolving Skills
- 2026EFLUX: Elastic Multi-Robot Formation Navigation and Adaptation with Agentic LLMs
- 2025MultiAgentBench: Evaluating the Collaboration and Competition of LLM agents
- 2025Multi-Agent Systems for Robotic Autonomy with LLMs
- 2025LLM Collaboration With Multi-Agent Reinforcement Learning
- 2025Natural Selection via Foundation Models for Soft Robot Evolution