Agents that Matter: How Subtle Choices Shape Agent Attribution in Multi-Agent Systems
cs.MA, cs.CL
Submitted: 2026-05-26
Updated: 2026-09-26
Code: https://github.com/ybkim95/agent-scaling
Terminology
Sources
- Program Synthesis with Large Language Models
- Model Cascading for Code: A Cascaded Black-Box Multi-Model Framework for Cost-Efficient Code Completion with Self-Testing
- MDTeamGPT: A Self-Evolving LLM-based Multi-Agent Framework for Multi-Disciplinary Team Medical Consultation
- BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- Efficient Leave-one-out Approximation in LLM Multi-agent Debate Based on Introspection
- Plancraft: an evaluation dataset for planning with LLM agents
- MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework
- Towards a Science of Scaling Agent Systems
- AgentBench: Evaluating LLMs as Agents
- A Dynamic LLM-Powered Agent Network for Task-Oriented Agent Collaboration
- RISE: Randomized Input Sampling for Explanation of Black-box Models
- OpenAI GPT-5 System Card
- WorkBench: a Benchmark Dataset for Agents in a Realistic Workplace Setting
- Qwen3 Technical Report
- MedEthicsQA: A Comprehensive Question Answering Benchmark for Medical Ethics Evaluation of LLMs
- Understanding and Optimizing Agentic Workflows via Shapley value
Related papers
- Highway Congestion Reduction through Reinforcement Learning Based Eulerian Headway Control
- 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
- StitchCUDA: An Automated Multi-Agents End-to-End GPU Programing Framework with Rubric-based Agentic Reinforcement Learning