EVOCHAMBER: Test-Time Co-evolution of Multi-Agent System at Individual, Team, and Population Scales

Hugging Face Daily Papers Papers

Summary

EVOCHAMBER is a training-free, multi-agent test-time evolution framework that enables emergent specialization through collaborative reflection and asymmetric knowledge transfer across individual, team, and population scales, achieving significant improvements on math, code, and reasoning tasks.

We argue that multi-agent test-time evolution is not single-agent evolution replicated N times. A single-agent learner can only evolve its own context and memory. A multi-agent system additionally evolves who collaborates, how they collaborate, and how knowledge flows across the population. These components have no single-agent counterpart and can produce phenomena such as emergent specialization. Yet prior test-time methods either confine experiences to individual agents, forfeiting cross-agent learning, or broadcast symmetrically to all agents, erasing the specialization that makes collaboration valuable. We present EVOCHAMBER, a training-free framework that instantiates test-time evolution at three levels over a coevolving agent pool. At its core is CODREAM (Collaborative Dreaming), a post-task protocol triggered on team failure or disagreement, in which agents collaboratively reflect, distill insights, and route them asymmetrically from strong to weak agents on the failed niche, preserving specialization while filling knowledge gaps. Team-level operators assemble niche-conditioned teams and select collaboration structures online. Population-level lifecycle operators fork, merge, prune, and seed agents under performance pressure. On three heterogeneous task streams with Qwen3-8B, EVOCHAMBER reaches 63.9% on competition math, 75.7% on code, and 87.1% on multi-domain reasoning, outperforming the best baseline by 32% relative on math and confirming asymmetric cross-agent transfer as the primary driver in ablation. Starting from several identically initialized agents, four to five stable niche specialists spontaneously emerge, a structural signature of multi-agent evolution that no single-agent learner can express. See our code at: https://github.com/Mercury7353/EvoChamber
Original Article
View Cached Full Text

Cached at: 05/14/26, 12:15 AM

Paper page - EVOCHAMBER: Test-Time Co-evolution of Multi-Agent System at Individual, Team, and Population Scales

Source: https://huggingface.co/papers/2605.11136

Abstract

Multi-agent test-time evolution framework EVOCHAMBER enables emergent specialization through collaborative reflection and asymmetric knowledge transfer across coevolving agents.

We argue thatmulti-agent test-time evolutionis notsingle-agent evolutionreplicated N times. A single-agent learner can only evolve its own context and memory. A multi-agent system additionally evolves who collaborates, how they collaborate, and how knowledge flows across the population. These components have no single-agent counterpart and can produce phenomena such asemergent specialization. Yet prior test-time methods either confine experiences to individual agents, forfeiting cross-agent learning, or broadcast symmetrically to all agents, erasing the specialization that makes collaboration valuable. We present EVOCHAMBER, a training-free framework that instantiates test-time evolution at three levels over acoevolving agent pool. At its core isCODREAM(Collaborative Dreaming), a post-task protocol triggered on team failure or disagreement, in which agents collaboratively reflect, distill insights, and route them asymmetrically from strong to weak agents on the failed niche, preserving specialization while filling knowledge gaps. Team-level operators assembleniche-conditioned teamsand selectcollaboration structuresonline.Population-level lifecycle operatorsfork, merge, prune, and seed agents under performance pressure. On three heterogeneous task streams with Qwen3-8B, EVOCHAMBER reaches 63.9% on competition math, 75.7% on code, and 87.1% on multi-domain reasoning, outperforming the best baseline by 32% relative on math and confirming asymmetric cross-agent transfer as the primary driver in ablation. Starting from several identically initialized agents, four to five stable niche specialists spontaneously emerge, a structural signature of multi-agent evolution that no single-agent learner can express. See our code at: https://github.com/Mercury7353/EvoChamber

View arXiv pageView PDFGitHub1Add to collection

Get this paper in your agent:

hf papers read 2605\.11136

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.11136 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.11136 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.11136 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

EvoTest: Evolutionary Test-Time Learning for Self-Improving Agentic Systems

arXiv cs.CL

EvoTest introduces J-TTL, a benchmark for measuring agent test-time learning capabilities, and proposes an evolutionary framework where an Actor Agent plays games while an Evolver Agent iteratively improves the system's prompts, memory, and hyperparameters without fine-tuning. The method demonstrates superior performance compared to reflection and memory-based baselines on complex text-based games.

EvoScientist: Towards Multi-Agent Evolving AI Scientists for End-to-End Scientific Discovery

Papers with Code Trending

EvoScientist is an adaptive multi-agent framework for end-to-end scientific discovery that continuously improves through persistent memory modules, comprising three specialized agents for idea generation, experiment execution, and knowledge distillation. It outperforms 7 state-of-the-art systems in scientific idea generation and improves code execution success rates through multi-agent evolution.