EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignment
cs.AI
Submitted: 2026-09-29
Updated: 2026-09-29
Code: https://github.com/beita6969/evosteer
Terminology
Sources
- CausalFlow: Causal Attribution and Counterfactual Repair for LLM Agent Failures
- SkillCAT: Contrastive, Assessment-Augmented and Topology-AwareSkill Self-Evolution for LLM Agents
- COLLATOR: Compositional Multi-Agent Orchestration with Counterfactual Reinforcement Learning
- The Trace Is the State: Exact Credit Assignment for LLM Agent Teams
- COSAC: Counterfactual Credit Assignment in Sequential Cooperative Teams
- $f$-Trajectory Balance: A Loss Family for Tuning GFlowNets, Generative Models, and LLMs with Off- and On-Policy Data
- SelfEvoSkill: Paired Execution Audits for Skill Revision without External Outcome Supervision
- Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systems
- SkillCommit: Evolving Agent Skills through Behaviorally Validated Scope Expansion
- Measuring Mathematical Problem Solving With the MATH Dataset
- SkillBrew: Multi-Objective Curation of Skill Banks for LLM Agents
- Don't Solve, Just Compare: Tiny Advisors for Runtime Intervention in LLM Agents
- Multi-Agent Coordination Adaptation via Structure-Guided Orchestration
- Experience as a Compass: Multi-agent RAG with Evolving Orchestration and Agent Prompts
- Counterfactual Credit Policy Optimization for Multi-Agent Collaboration
- Last Step Matters: Early Uncertainty Cannot Predict Failure in Long-Horizon Agents
- Granularity-Adaptive Credit Assignment for Long-Horizon LLM Agent Reinforcement Learning
- GFlowRL: Scaling Distribution-Matching RL to Large Language Models
- Autonomous Repair for Multi-Agent Systems via Monte-Carlo Tree Search
- Repair or Resample? Rethinking Failure Debugging in LLM Multi-Agent Systems
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems