MedVerse: Efficient and Reliable Medical Reasoning via DAG-Structured Parallel Execution

TL;DR

MedVerse introduces a DAG-structured parallel reasoning framework for medical inference, enhancing efficiency and reliability of large language models in complex medical tasks.

Contribution

It reformulates medical reasoning as a DAG process, incorporating a new data pipeline, topology-aware attention, and a parallel inference engine, with code available online.

Findings

Up to 8.9% performance improvement over strong LLMs.

1.3x reduction in inference latency compared to specialized medical LLMs.

1.7x increase in generation throughput due to parallel decoding.

Abstract

Large language models (LLMs) have demonstrated strong performance and rapid progress in a wide range of medical reasoning tasks. However, their sequential autoregressive decoding forces inherently parallel clinical reasoning, such as differential diagnosis, into a single linear reasoning path, limiting both efficiency and reliability for complex medical problems. To address this, we propose MedVerse, a reasoning framework for complex medical inference that reformulates medical reasoning as a parallelizable directed acyclic graph (DAG) process based on Petri net theory. The framework adopts a full-stack design across data, model architecture, and system execution. For data creation, we introduce the MedVerse Curator, an automated pipeline that synthesizes knowledge-grounded medical reasoning paths and transforms them into Petri net-structured representations. At the architectural level, we propose a topology-aware attention mechanism with adaptive position indices that supports parallel reasoning while preserving logical consistency. Systematically, we develop a customized inference engine that supports parallel execution without additional overhead. Empirical evaluations show that MedVerse improves strong general-purpose LLMs by up to 8.9%. Compared to specialized medical LLMs, MedVerse achieves comparable performance while delivering a 1.3x reduction in inference latency and a 1.7x increase in generation throughput, enabled by its parallel decoding capability. Code is available at https://github.com/aiming-lab/MedVerse.