MedZERO: Self-Evolving Agents for Open-Ended Medical Reasoning Through Controlled Knowledge Accumulation

Key points

• Large language models (LLMs) have shown promise in medical question answering and clinical reasoning, yet their improvement remains constrained by static parametric knowledge and costly expert supervision.1
• Self-evolving agents offer a promising alternative by enabling models to improve through iterative task generation and problem-solving.1
• MedZERO couples an Examiner that generates frontier medical question-option pairs with a Reasoner that solves them through evidence-grounded multi-turn reasoning with external knowledge tools.1
• We evaluate MedZERO on five public medical reasoning benchmarks using 4B- and 8B-scale base models under open-ended evaluation.1

Sources (1)

• [1]MedZERO: Self-Evolving Agents for Open-Ended Medical Reasoning Through Controlled Knowledge Accumulation
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 01:25 PM

  We present MedZERO, a self-evolving framework for open-ended medical reasoning.

  Large language models (LLMs) have shown promise in medical question answering and clinical reasoning, yet their improvement remains constrained by static parametric knowledge and costly expert supervision.

Extractive summary: sentences quoted from the sources.