mlsanityData to 7 Oct 2026

CERA-MoA: Co-Evolving Routing Mechanisms with Continually Learning LLM Agents

arXiv 2609.18779 ↗First posted 16 Sep 2026cs.AI cs.LGcited by 0

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

Current Mixture-of-Agents (MoA) paradigms generally treat query routing and agent fine-tuning as separate processes, limiting their ability to respond to evolving agent capabilities. This disconnect prevents routing strategies from adapting to evolving agent capabilities during post-training and prevents agents from achieving synergistic data-driven specialization. To resolve this, we introduce CERA-MoA (Co-Evolving Router with continually learning Agents for Mixture-of-Agents), an iterative reinforcement learning framework where the dynamic router and independent agent policies co-evolve. We design a predictive familiarity estimator that leverages mid-layer hidden states to evaluate semantic competence among agents, avoiding the overhead of full rollouts. Based on these familiarity scores, a cumulative-threshold adaptive routing mechanism dynamically activates a tailored minimal agent subset, achieving a trade-off between task performance and efficiency. By proactively allocating targeted training samples to agents based on their evolving competence, CERA-MoA promotes capability differentiation. Extensive experiments across various domains demonstrate that CERA-MoA outperforms state-of-the-art static-agent routing and fix-workflow fine-tuning baselines.

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 50 of its 59 references that are in the map. First posted on arXiv 16 Sep 2026.

Builds on

50 of 59 references.

Closest ancestors · 10

Same direction, newest first

  1. 2026RouteMoA: Dynamic Routing without Pre-Inference Boosts Efficient Mixture-of-Agents26 Jan
  2. 2026OFA-MAS: One-for-All Multi-Agent System Topology Design based on Mixture-of-Experts Graph Generative Models19 Jan
  3. 2026Orchestrating Intelligence: Confidence-Aware Routing for Efficient Multi-Agent Collaboration across Multi-Scale Models8 Jan
  4. 2026EvoRoute: Experience-Driven Self-Routing LLM Agent Systems6 Jan
  5. 2025CoMAS: Co-Evolving Multi-Agent Systems via Interaction Rewards9 Oct
  6. 2025Reinforce LLM Reasoning through Multi-Agent Reflection10 Jun
  7. 2025RMoA: Optimizing Mixture-of-Agents through Diversity Maximization and Residual Compensation30 May
  8. 2025Multi-Agent Collaboration via Evolving Orchestration26 May
  9. 2024MALT: Improving Reasoning with Multi-Agent LLM Training2 Dec
  10. 2024Mixture-of-Agents Enhances Large Language Model Capabilities7 Jun

Other references · 40

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  3. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  4. 2025Qwen3 Technical Reportcited by 8,195
  5. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  6. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  7. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,710
  8. 2021Program Synthesis with Large Language Modelscited by 3,509
  9. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,678
  10. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,767
  11. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,143
  12. 2024OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problemscited by 1,099
28 more
  1. 2023Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debatecited by 995
  2. 2024Tulu 3: Pushing Frontiers in Open Language Model Post-Trainingcited by 988
  3. 2025Group Sequence Policy Optimizationcited by 788
  4. 2023AGIEval: A Human-Centric Benchmark for Evaluating Foundation Modelscited by 701
  5. 2024Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothingcited by 341
  6. 2023ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMscited by 297
  7. 2025Layer by Layer: Uncovering Hidden Representations in Language Modelscited by 293
  8. 2025Learning to Reason without External Rewardscited by 199
  9. 2024Free Process Rewards without Process Labelscited by 162
  10. 2023TACO: Topics in Algorithmic COde generation datasetcited by 150
  11. 2025Confidence Improves Self-Consistency in LLMscited by 111
  12. 2025MAPoRL: Multi-Agent Post-Co-Training for Collaborative Large Language Models with Reinforcement Learningcited by 80
  13. 2024LogicBench: Towards Systematic Evaluation of Logical Reasoning Ability of Large Language Modelscited by 67
  14. 2025ReVISE: Learning to Refine at Test-Time via Intrinsic Self-Verificationcited by 35
  15. 2024Multilinear Mixture of Experts: Scalable Expert Specialization through Factorizationcited by 28
  16. 2024Which LLM to Play? Convergence-Aware Online Model Selection with Time-Increasing Banditscited by 26
  17. 2025ICL-Router: In-Context Learned Model Representations for LLM Routingcited by 26
  18. 2026MAS-Orchestra: Understanding and Improving Multi-Agent Reasoning Through Holistic Orchestration and Controlled Benchmarkscited by 26
  19. 2025Stronger-MAS: Multi-Agent Reinforcement Learning for Collaborative LLMscited by 25
  20. 2025From Debate to Equilibrium: Belief-Driven Multi-Agent LLM Reasoning via Bayesian Nash Equilibriumcited by 24
  21. 2025Multi-Agent Debate for LLM Judges with Adaptive Stability Detectioncited by 20
  22. 2025Online Multi-LLM Selection via Contextual Bandits under Unstructured Context Evolutioncited by 18
  23. 2024Fine-Tuning on Diverse Reasoning Chains Drives Within-Inference CoT Refinement in LLMscited by 12
  24. 2025LD-MoLE: Learnable Dynamic Routing for Mixture of LoRA Expertscited by 12
  25. 2025Coupling Experts and Routers in Mixture-of-Experts via an Auxiliary Losscited by 11
  26. 2025iMAD: Intelligent Multi-Agent Debate for Efficient and Accurate LLM Inferencecited by 10
  27. 2025Confidence-Guided Stepwise Model Routing for Cost-Efficient Reasoningcited by 9
  28. 2026Epistemic Gain, Aleatoric Cost: Uncertainty Decomposition in Multi-Agent Debate for Math Reasoningcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026MonoScale: Scaling Multi-Agent System with Monotonic Improvement
  2. 2026HM-ROUTER: Joint Model and Harness Routing for Agentic Systems
  3. 2026MMoA: An AI-Agent framework with recurrence for Memoried Mixure-of-Agent
  4. 2026MetaClaw: Just Talk -- An Agent That Meta-Learns and Evolves in the Wild
  5. 2026MERA: Model Evolution and Routing with Skill Adaptation for Agentic Systems at Scale
  6. 2026Structuring MoE Expert Selection for Agentic Reinforcement Learning
  7. 2025Adaptive Minds: Empowering Agents with LoRA-as-Tools
  8. 2026ARCO: Adaptive Rubrics with Co-Evolution for Multi-Step LLM-Based Agents