{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T06:46:29Z","timestamp":1785912389047,"version":"3.56.0"},"reference-count":57,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,3]]},"DOI":"10.1109\/wacvw68408.2026.00036","type":"proceedings-article","created":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T19:18:34Z","timestamp":1785871114000},"page":"303-313","source":"Crossref","is-referenced-by-count":0,"title":["What Matters in Reinforcement Learning Based Training for Medical Vision Language Model: an Empirical Study"],"prefix":"10.1109","author":[{"given":"Wenhui","family":"Zhu","sequence":"first","affiliation":[{"name":"Arizona State University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xuanzhao","family":"Dong","sequence":"additional","affiliation":[{"name":"Arizona State University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xin","family":"Li","sequence":"additional","affiliation":[{"name":"Arizona State University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peijie","family":"Qiu","sequence":"additional","affiliation":[{"name":"Washington University in St. Louis"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiwen","family":"Chen","sequence":"additional","affiliation":[{"name":"Clemson University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Walker","family":"Northrup","sequence":"additional","affiliation":[{"name":"Arizona State University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Abolfazl","family":"Razi","sequence":"additional","affiliation":[{"name":"Clemson University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aris","family":"Sotiras","sequence":"additional","affiliation":[{"name":"Washington University in St. Louis"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yi","family":"Su","sequence":"additional","affiliation":[{"name":"Banner Alzheimer's Institute"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhipeng","family":"Wang","sequence":"additional","affiliation":[{"name":"LinkedIn Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shao","family":"Tang","sequence":"additional","affiliation":[{"name":"LinkedIn Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yalin","family":"Wang","sequence":"additional","affiliation":[{"name":"Arizona State University"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Cosmos-reason1: From physical common sense to embodied reasoning","author":"Azzolini","year":"2025","journal-title":"arXiv preprint"},{"key":"ref2","article-title":"Cosmos-reason1: From physical common sense to embodied reasoning","author":"Azzolini","year":"2025","journal-title":"arXiv preprint"},{"key":"ref3","article-title":"Qwen2. 5-vl technical report","author":"Bai","year":"2025","journal-title":"arXiv preprint"},{"key":"ref4","first-page":"1877","article-title":"Language models are few-shot learners","volume":"33","author":"Brown","year":"2020","journal-title":"Advances in neural information processing systems"},{"key":"ref5","volume-title":"R1-v: Reinforcing super generalization ability in visionlanguage models with less than 3","author":"Chen","year":"2025"},{"key":"ref6","article-title":"Chexagent: Towards a foundation model for chest x-ray interpretation","author":"Chen","year":"2024","journal-title":"arXiv preprint"},{"key":"ref7","article-title":"Vision-language models can self-improve reasoning via reflection","author":"Cheng","year":"2024","journal-title":"arXiv preprint"},{"issue":"70","key":"ref8","first-page":"1","article-title":"Scaling instructionfinetuned language models","volume":"25","author":"Won Chung","year":"2024","journal-title":"Journal of Machine Learning Research"},{"key":"ref9","article-title":"aha!","author":"Chung","year":"2025","journal-title":"moment"},{"key":"ref10","article-title":"Llada-medv: Exploring large language diffusion models for biomedical image understanding","author":"Dong","year":"2025","journal-title":"arXiv preprint"},{"key":"ref11","article-title":"Insight-v: Exploring long-chain visual reasoning with multimodal large language models","author":"Dong","year":"2024","journal-title":"arXiv preprint"},{"key":"ref12","article-title":"The llama 3 herd of models","author":"Grattafiori","year":"2024","journal-title":"arXiv preprint"},{"key":"ref13","article-title":"Deepseek-r1: Incentivizing reasoning capability in 1lms via reinforcement learning","author":"Guo","year":"2025","journal-title":"arXiv preprint"},{"issue":"2","key":"ref14","first-page":"3","article-title":"Lora: Low-rank adaptation of large language models","volume":"1","author":"Edward","year":"2022","journal-title":"ICLR"},{"key":"ref15","article-title":"Vision-r1: Incentivizing reasoning capability in multimodal large language models","author":"Huang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref16","article-title":"Vision-r1: Incentivizing reasoning capability in multimodal large language models","author":"Huang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref17","article-title":"Gpt-4o system card","author":"Hurst","year":"2024","journal-title":"arXiv preprint"},{"key":"ref18","volume-title":"Mistral 7b","author":"Albert","year":"2023"},{"key":"ref19","article-title":"Llm post-training: A deep dive into reasoning large language models","author":"Kumar","year":"2025","journal-title":"arXiv preprint"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-acl.348"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/tmi.2026.3661001"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/tmi.2026.3661001"},{"key":"ref23","volume-title":"Qwen2-vl-finetune","author":"Lee","year":"2024"},{"key":"ref24","first-page":"19730","article-title":"Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models","volume-title":"International conference on machine learning","author":"Li","year":"2023"},{"key":"ref25","article-title":"Llava-surg: towards multimodal surgical assistant via structured surgical video learning","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref27","article-title":"Understanding r1-zero-like training: A critical perspective","author":"Liu","year":"2025","journal-title":"arXiv preprint"},{"key":"ref28","article-title":"INGIN LLMS","year":"2025","journal-title":"Demystifying long chain-of-thought reason"},{"key":"ref29","article-title":"Ui-r1: Enhancing action prediction of gui agents by reinforcement learning","author":"Lu","year":"2025","journal-title":"arXiv preprint"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbac409"},{"key":"ref31","article-title":"Biomedgpt: Open multimodal generative pre-trained transformer for biomedicine","author":"Luo","year":"2023","journal-title":"arXiv preprint"},{"key":"ref32","first-page":"353","article-title":"Med-flamingo: a multimodal medical few-shot learner","author":"Moor","year":"2023","journal-title":"Machine Learning for Health (ML4H)"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-032-04981-0_32"},{"key":"ref34","article-title":"Scaling up biomedical visionlanguage models: Fine-tuning, instruction tuning, and multimodal learning","author":"Peng","year":"2025","journal-title":"arXiv preprint"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2338"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1410"},{"key":"ref37","volume-title":"Proximal policy optimization algorithms","author":"Schulman","year":"2017"},{"key":"ref38","article-title":"Deepseekmath: Pushing the limits of mathematical reasoning in open language models","author":"Shao","year":"2024","journal-title":"arXiv preprint"},{"key":"ref39","article-title":"Vlm-r1: A stable and generalizable r1-style large vision-language model","author":"Shen","year":"2025","journal-title":"arXiv preprint"},{"key":"ref40","article-title":"Vlmr1: A stable and generalizable r1-style large vision-language model","author":"Shen","year":"2025","journal-title":"arXiv preprint"},{"key":"ref41","article-title":"Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context","author":"Team","year":"2024","journal-title":"arXiv preprint"},{"key":"ref42","article-title":"Kimi k1. 5: Scaling reinforcement learning with 1lms","author":"Team","year":"2025","journal-title":"arXiv preprint"},{"key":"ref43","article-title":"Endochat: Grounded multimodal large language model for endoscopic surgery","author":"Wang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref44","volume-title":"Qwen2-vl: Enhancing vision-language model\u2019s perception of the world at any resolution","author":"Wang","year":"2024"},{"key":"ref45","article-title":"Careaqa: A cardiac and respiratory audio question answering model for open-ended diagnostic reasoning","author":"Wang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref46","volume-title":"Open-r1-video","author":"Wang","year":"2025"},{"key":"ref47","article-title":"Gui-r1: A generalist r1-style vision-language action model for gui agents","author":"Xia","year":"2025","journal-title":"arXiv preprint"},{"key":"ref48","first-page":"arXiv-2305","article-title":"Biomedgpt: A unified and generalist biomedical generative pre-trained transformer for vision, language, and multimodal tasks","author":"Zhang","year":"2023","journal-title":"arXiv e-prints"},{"key":"ref49","article-title":"Improve vision language model chain-ofthought reasoning","author":"Zhang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref50","article-title":"Instruction tuning for large language models: A survey","author":"Zhang","year":"2023","journal-title":"arXiv preprint"},{"key":"ref51","article-title":"Pmc-vqa: Visual instruction tuning for medical visual question answering","author":"Zhang","year":"2023","journal-title":"arXiv preprint"},{"key":"ref52","article-title":"Respllm: Unifying audio and text with multimodal 11ms for generalized respiratory health prediction","author":"Zhang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref53","article-title":"R1-omni: Explainable omni-multimodal emotion recognition with reinforcement learning","author":"Zhao","year":"2025","journal-title":"arXiv preprint"},{"key":"ref54","volume-title":"Easyr1: An efficient, scalable, multi-modality rl training framework","author":"Zheng","year":"2025"},{"key":"ref55","article-title":"aha moment","author":"Zhou","year":"2025","journal-title":"visual reasoning on a 2b non-sft model"},{"key":"ref56","article-title":"R1-zero\u2019s\u201d aha moment\u201d in visual reasoning on a 2b non-sft model","author":"Zhou","year":"2025","journal-title":"arXiv preprint"},{"key":"ref57","article-title":"Retinalgpt: A retinal clinical preference conversational assistant powered by large visionlanguage models","author":"Zhu","year":"2025","journal-title":"arXiv preprint"}],"event":{"name":"2026 IEEE\/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)","location":"Tucson, AZ, USA","start":{"date-parts":[[2026,3,6]]},"end":{"date-parts":[[2026,3,10]]}},"container-title":["2026 IEEE\/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11617369\/11617396\/11617819.pdf?arnumber=11617819","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T05:59:26Z","timestamp":1785909566000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11617819\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3]]},"references-count":57,"URL":"https:\/\/doi.org\/10.1109\/wacvw68408.2026.00036","relation":{},"subject":[],"published":{"date-parts":[[2026,3]]}}}