EARS: Explanatory Abstention for Reliable Sub-Agent Modeling in Large-scale Multi-Agent Systems

arXiv:2606.18668 · cs.MA, cs.CL · Submitted 2026-06-17 · Read on arXiv

cs.MA, cs.CL

Submitted: 2026-06-17

Updated: 2026-10-09

Terminology

• Abstention — 5× in this paper · explained in 1 episode
• Explanatory Abstention for Reliable Sub-Agent Modeling — 2× in this paper
• Multi-Agent System — 2× in this paper · explained in 6 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• LLM-as-a-Judge — 1× in this paper · explained in 3 episodes
• Modularity — 1× in this paper · explained in 9 episodes

Sources

• Phi-4 Technical Report
• GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning
• Knowledge of Knowledge: Exploring Known-Unknowns Uncertainty with Large Language Models
• Why Do Multi-Agent LLM Systems Fail?
• Don't Hallucinate, Abstain: Identifying LLM Knowledge Gaps via Multi-LLM Collaboration
• Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks
• Accelerating scientific discovery with Co-Scientist
• Controlling Performance and Budget of a Centralized Multi-agent LLM System with Reinforcement Learning
• AbstentionBench: Reasoning LLMs Fail on Unanswerable Questions
• MediQ: Question-Asking LLMs and a Benchmark for Reliable Interactive Clinical Reasoning
• Qwen3 Technical Report
• LatentRefusal: Latent-Signal Refusal for Unanswerable Text-to-SQL Queries
• Replacing Judges with Juries: Evaluating LLM Generations with a Panel of Diverse Models
• OpenHands: An Open Platform for AI Software Developers as Generalist Agents
• Know Your Limits: A Survey of Abstention in Large Language Models
• Talk to Right Specialists: Iterative Routing in Multi-agent Systems for Question Answering
• AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
• MasRouter: Learning to Route LLMs for Multi-Agent Systems
• Abstain-R1: Calibrated Abstention and Post-Refusal Clarification via Verifiable RL
• AgentOrchestra: Orchestrating Multi-Agent Intelligence with the Tool-Environment-Agent(TEA) Protocol

Related papers

• Highway Congestion Reduction through Reinforcement Learning Based Eulerian Headway Control
• ORACLE-SWE: Quantifying the Contribution of Critical Information Signals on SWE Agents
• You Only Align Once: Propagating Cooperative Behaviors in Multi-Agent Systems through Seed Agents
• Deny Without Disabling: Authorization-Paired Evaluation and Control for Multi-Agent Systems
• MA-SAPO: Multi-Agent Reasoning for Score-Aware Prompt Optimization
• PeroMAS: A Multi-agent System of Perovskite Material Discovery