# MOSAIC: A Universal Agent-Level Interface for Cross-Paradigm Agent Mixing and Human-AI Collaboration

arXiv: https://arxiv.org/abs/2603.01260 · HTML: https://mlsanity.com/paper/2603.01260/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use.
- Builds on 20 of its 26 references that are in the map.
- First posted on arXiv 1 Mar 2026.

## Placed in

- LLM agents and tool use / [Reinforcement learning for LLM agents](https://mlsanity.com/agents/reinforcement-learning-for-llm-agents/)

## Builds on (most cited first, 20 of 20)

- [On the Utility of Learning about Humans for Human-AI Coordination](https://mlsanity.com/paper/1910.05789/) (2019)
- [PettingZoo: Gym for Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2009.14471/) (2020)
- [Tianshou: a Highly Modularized Deep Reinforcement Learning Library](https://mlsanity.com/paper/2107.14171/) (2021)
- [GameBench: Evaluating Strategic Reasoning Abilities of LLM Agents](https://mlsanity.com/paper/2406.06613/) (2024)
- [Sample Factory: Egocentric 3D Control from Pixels at 100000 FPS with Asynchronous Reinforcement Learning](https://mlsanity.com/paper/2006.11751/) (2020)
- [JaxMARL: Multi-Agent RL Environments and Algorithms in JAX](https://mlsanity.com/paper/2311.10090/) (2023)
- [EnvPool: A Highly Parallel Reinforcement Learning Environment Execution Engine](https://mlsanity.com/paper/2206.10558/) (2022)
- [TorchRL: A data-driven decision-making library for PyTorch](https://mlsanity.com/paper/2306.00577/) (2023)
- [lmgame-Bench: How Good are LLMs at Playing Games?](https://mlsanity.com/paper/2505.15146/) (2025)
- [A Survey of Ad Hoc Teamwork Research](https://mlsanity.com/paper/2202.10450/) (2022)
- [Collab-Overcooked: Benchmarking and Evaluating Large Language Models as Collaborative Agents](https://mlsanity.com/paper/2502.20073/) (2025)
- [BenchMARL: Benchmarking Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2312.01472/) (2023)
- [ZSC-Eval: An Evaluation Toolkit and Benchmark for Multi-agent Zero-shot Coordination](https://mlsanity.com/paper/2310.05208/) (2023)
- [Evaluating Large Language Models with Grid-Based Game Competitions: An Extensible LLM Benchmark and Leaderboard](https://mlsanity.com/paper/2407.07796/) (2024)
- [GAMEBoT: Transparent Assessment of LLM Reasoning in Games](https://mlsanity.com/paper/2412.13602/) (2024)
- [N-Agent Ad Hoc Teamwork](https://mlsanity.com/paper/2404.10740/) (2024)
- [LLM CHESS: Benchmarking Reasoning and Instruction-Following in LLMs through Chess](https://mlsanity.com/paper/2512.01992/) (2025)
- [A General Learning Framework for Open Ad Hoc Teamwork Using Graph-based Policy Learning](https://mlsanity.com/paper/2210.05448/) (2022)
- [XuanCe: A Comprehensive and Unified Deep Reinforcement Learning Library](https://mlsanity.com/paper/2312.16248/) (2023)
- [BotzoneBench: Scalable LLM Evaluation via Graded AI Anchors](https://mlsanity.com/paper/2602.13214/) (2026)

## Abstract (arXiv)

Existing infrastructure cannot deploy agents from different decision-making paradigms within the same environment, making fair cross-paradigm comparison under identical conditions impossible. We present MOSAIC, an open-source platform that enables heterogeneous agents (RL policies, LLMs, VLMs, and human operators) to act within shared reinforcement learning environments in ad-hoc team settings with reproducible results. MOSAIC introduces three contributions. (i) IPC-based worker protocol that wraps native and third-party frameworks as isolated subprocess workers, each executing its own training and inference logic unmodified and communicating through a versioned inter-process protocol. (ii) An operator abstraction that forms an agent-level interface by mapping workers to agent slots: each operator, regardless of whether it is backed by an RL policy, an LLM, or a human, conforms to a minimal universal interface. (iii) A deterministic cross-paradigm evaluation framework with two complementary modes: a manual mode that advances up to N operators in lock-step under shared seeds for fine-grained visual inspection of behavioural differences; and a script mode that drives automated, long-running evaluation via declarative Python scripts for reproducible experiments. Our documentation is released at: https://mosaic-platform.readthedocs.io.