Evaluating the Transfer of Co-Evolved Communication from 2D to 3D Simulation

Hugging Face Daily Papers Papers

Summary

该论文研究两个机器人智能体在二维离散环境中共同演化出的通信机制,在迁移到基于物理的三维模拟器后是否仍能保持其功能作用。结果显示通信信号主要作为辅助解堵机制而非导航引导,且更长的评估时长能持续提升性能。

This work examines the transfer of a co-evolved communication mechanism between two robotic agents from a discrete two-dimensional (2D) simulator to a three-dimensional simulator with real physics (3D). The study focuses on whether a communication mechanism co-evolved in a 2D environment retains its functional role after transfer to a 3D physics-based simulator. To support this analysis, the effects of the episode time budget, the social cue, and the asymmetry between the two co-evolved roles were examined. The results indicate that the success rate increased approximately linearly with the evaluated time budgets, with no evidence of a plateau between 2,000 and 6,000 physics steps, suggesting that evaluations based on shorter episodes may underestimate the performance of the trained controllers. In both simulators, the social cue functioned primarily as a jam- assistance mechanism rather than as a navigation guide, although with a more pronounced effect in 2D. Analysis of eight independent evolutionary runs revealed a consistent direction of asymmetry, although its magnitude varied across runs. Controlling the processing order between agents allowed us to rule out an artifact of the physics engine. Finally, the results are discussed in terms of the factors that may contribute to the remaining performance gap observed after transfer.
Original Article
View Cached Full Text

Cached at: 10/10/26, 02:24 AM

Paper page - Evaluating the Transfer of Co-Evolved Communication from 2D to 3D Simulation

Source: https://huggingface.co/papers/2610.09280 This paper investigates whether a communication mechanism co-evolved between two robotic agents in a discrete 2D environment preserves its functional role after transfer to a 3D physics-based simulator. We analyze the effects of episode duration, social cues, and role asymmetry across multiple independent evolutionary runs.

Results show that performance continues to improve with longer evaluation horizons, suggesting that short episodes may underestimate learned behaviors. In both simulators, the communication signal acts primarily as a jam-assistance mechanism rather than a navigation guide, although the effect is stronger in the original 2D environment.

These findings contribute to understanding how emergent communication and cooperative behaviors transfer across simulation fidelities, an important challenge in evolutionary robotics and sim-to-sim transfer.

Similar Articles

CoDA: Co-Evolution of Data Engines and Action Policies

Hugging Face Daily Papers

CoDA 提出一种仿真数据引擎与机器人动作策略协同演进的框架,通过共享失败与学习进展记忆、视觉语言诊断以及 Text2potential 分阶段势能奖励,将机器人操作任务的分布外成功率从 5.3% 提升至 27.6%,并在 LIBERO-Pro 上取得显著改善。

Words to World: Benchmarking LLMs in Interactive Physical Environments

Hugging Face Daily Papers

该论文提出了一个用于评估大语言模型在闭环物理交互中表现的基准,模型需通过连接机械臂、摄像头和传感器的系统在真实环境中执行任务并根据物理反馈调整行动。实验揭示了LLM在物理行动中的常见失败模式,如幻觉状态变化、糟糕的异常恢复和不稳定的长程规划。