Med-RLVR: Emerging Medical Reasoning from a 3B base model via reinforcement Learning
This demonstrates RLVR's applicability beyond math/coding to medical QA, potentially benefiting AI-assisted healthcare, though it's an incremental extension of an existing method to a new domain.
The authors investigated whether reinforcement learning from verifiable rewards (RLVR) could elicit medical reasoning from a 3B-parameter base model without explicit supervision, using medical multiple-choice question answering data. They found RLVR achieved comparable in-distribution performance to supervised fine-tuning while improving out-of-distribution generalization by 8 accuracy points.
Reinforcement learning from verifiable rewards (RLVR) has recently gained attention for its ability to elicit self-evolved reasoning capabilitie from base language models without explicit reasoning supervisions, as demonstrated by DeepSeek-R1. While prior work on RLVR has primarily focused on mathematical and coding domains, its applicability to other tasks and domains remains unexplored. In this work, we investigate whether medical reasoning can emerge from RLVR. We introduce Med-RLVR as an initial study of RLVR in the medical domain leveraging medical multiple-choice question answering (MCQA) data as verifiable labels. Our results demonstrate that RLVR is not only effective for math and coding but also extends successfully to medical question answering. Notably, Med-RLVR achieves performance comparable to traditional supervised fine-tuning (SFT) on in-distribution tasks while significantly improving out-of-distribution generalization, with an 8-point accuracy gain. Further analysis of training dynamics reveals that, with no explicit reasoning supervision, reasoning emerges from the 3B-parameter base model. These findings underscore the potential of RLVR in domains beyond math and coding, opening new avenues for its application in knowledge-intensive fields such as medicine.