Meta tags:
author=Yunfei Xie;

Headings (most frequently used words):

news, selected, publications, education, experience, reviewer,

Text of the page (most frequently used words):
and (21), xie (16), for (15), 2026 (10), 2025 (10), segmentation (9), zhou (9), with (9), yunfei (9), prof (8), university (8), research (8), the (8), 2024 (7), iclr (7), cihang (7), paper (7), mei (6), yuyin (6), icml (5), jieru (5), science (5), that (5), play (5), website (4), from (4), multimodal (4), models (4), part (4), global (4), training (4), story (4), chen (4), wei (4), accepted (4), cvpr (3), isbi (3), eccv (3), multi (3), medical (3), 2023 (3), dec (3), intern (3), language (3), medicine (3), computer (3), aug (3), rice (3), through (3), code (3), model (3), local (3), context (3), objects (3), aggregation (3), object (3), alan (3), yuille (3), framework (3), memo (3), games (3), learning (3), generalize (3), this (2), was (2), neurips (2), supervisor (2), focus (2), johns (2), hopkins (2), jul (2), ccvl (2), santa (2), cruz (2), feb (2), vlaa (2), artificial (2), intelligence (2), huazhong (2), technology (2), developed (2), two (2), transformer (2), brain (2), tumor (2), xianhang (2), clinical (2), preliminary (2), reasoning (2), data (2), juncheng (2), are (2), hierarchical (2), based (2), performance (2), pixels (2), free (2), long (2), iterative (2), text (2), more (2), adapter (2), multigranular (2), annotations (2), medtrinity (2), 25m (2), memory (2), run (2), wang (2), llm (2), reinforcement (2), during (2), lambda (2), open (2), template, borrowed, last, updated, march, 16th, jon, barron, ieee, iccv, tmm, tpami, reviewer, level, few, shot, california, experience, eng, sep, 2021, jun, wuhan, china, present, houston, education, cnn, hybrid, brats, challenge, create, broad, applicability, supervoxel, introduce, comprehensive, benchmark, evaluate, openai, capabilities, report, results, indicating, stronger, together, known, limitations, actionable, directions, improvement, arxiv, qiao, jin, yongshuo, zong, bingchen, zhao, siwei, yang, haoqin, study, closer, doctor, superpixel, simultaneously, addresses, conflicting, needs, detail, parts, employing, superpixels, group, tokens, achieved, state, art, approach, using, proposed, efficient, generating, stories, 100, frames, designed, paradigm, progressively, refines, process, integrating, constraints, achieving, precise, interactions, improved, semantic, consistency, throughout, bingjie, mude, hui, yuanqi, chang, xiaoke, huang, jiawei, mao, visualization, develop, rag, enhanced, automated, synthesis, pipeline, generates, million, high, quality, samples, vision, pretraining, james, zou, lei, xing, sheng, liu, hong, lang, gao, large, scale, dataset, propose, self, optimizes, inference, time, retention, exploration, raising, mean, win, rate, gpt, mini, across, five, while, reducing, variance, zhangyang, pramod, viswanath, cheston, tan, hongyuan, yihan, alexander, duffy, zhizhou, sha, jianzhu, yao, bobby, cheng, kevin, augmented, optimization, robust, turn, agent, show, post, simple, arcade, such, snake, improves, math, broader, benchmarks, without, any, exposure, worked, solutions, equations, diagrams, junfei, xiao, shiyi, lan, yinsong, reason, game, works, please, check, here, selected, publications, jan, started, journey, working, received, 000, thank, you, grant, presented, hosted, sentient, agi, symposium, excited, start, internship, summer, adobe, apr, may, news, photo, credit, partner, xuan, resume, twitter, github, google, scholar, email, collaborations, completed, bachelor, degree, school, automation, undergraduate, studies, honored, work, first, year, phd, student, department, advised, current, interests, lie, agents, 谢云飞,

Text of the page (random words):
yunfei xie yunfei xie 谢云飞 i m a first year cs phd student in the department of computer science at rice university advised by prof chen wei my current research interests lie in llm agents and reinforcement learning i completed my bachelor s degree in the school of artificial intelligence and automation at huazhong university of science technology during my undergraduate studies i was honored to work as a research intern with prof yuyin zhou and prof cihang xie in vlaa at uc santa cruz and with prof alan yuille and dr jieru mei in ccvl at johns hopkins university i am open for collaborations in research email google scholar github twitter resume photo credit my partner xuan news may 2026 memo is accepted to icml 2026 apr 2026 excited to start a research internship at adobe research this summer feb 2026 play to generalize and story adapter are accepted to iclr 2026 dec 2025 presented play to generalize at neurips 25 at the open agi symposium hosted by sentient aug 2025 received the lambda research grant 10 000 thank you lambda aug 2025 started my ph d journey in computer science at rice university working with prof chen wei jan 2025 medtrinity 25m is accepted to iclr 2025 jul 2024 from pixels to objects is accepted to eccv 2024 selected publications for more works please check here play to generalize learning to reason through game play yunfei xie yinsong ma shiyi lan alan yuille junfei xiao chen wei iclr 2026 paper website tl dr we show that post training multimodal language models with reinforcement learning on simple arcade games such as snake improves performance on math and broader reasoning benchmarks without any exposure to worked solutions equations or diagrams during training memo memory augmented model context optimization for robust multi turn multi agent llm games yunfei xie kevin wang bobby cheng jianzhu yao zhizhou sha alexander duffy yihan xi hongyuan mei cheston tan chen wei pramod viswanath zhangyang wang icml 2026 paper website code tl dr we propose memo a self play framework that optimizes inference time context through memory retention and exploration raising mean win rate from 24 9 to 49 5 for gpt 4o mini across five text based games while reducing run to run variance medtrinity 25m a large scale multimodal dataset with multigranular annotations for medicine yunfei xie ce zhou lang gao juncheng wu xianhang li hong yu zhou sheng liu lei xing james zou cihang xie yuyin zhou iclr 2025 paper website tl dr we develop a rag enhanced automated data synthesis pipeline that generates 25 million high quality multimodal medical samples with multigranular annotations for vision language model pretraining story adapter a training free iterative framework for long story visualization jiawei mao xiaoke huang yunfei xie yuanqi chang mude hui bingjie xu yuyin zhou iclr 2026 paper code tl dr we proposed a training free and efficient framework for generating long stories with up to 100 frames we designed an iterative paradigm that progressively refines the process by integrating text and global constraints achieving more precise interactions and improved semantic consistency throughout the story from pixels to objects a hierarchical approach for part and object segmentation using local and global aggregation yunfei xie cihang xie alan yuille jieru mei eccv 2024 paper tl dr we developed a hierarchical superpixel based model that simultaneously addresses two conflicting needs in segmentation local detail for parts and global context for objects by employing local aggregation for superpixels in part segmentation and global aggregation for group tokens in object segmentation we achieved state of the art performance in part and object segmentation a preliminary study of o1 in medicine are we closer to an ai doctor yunfei xie juncheng wu haoqin tu siwei yang bingchen zhao yongshuo zong qiao jin cihang xie yuyin zhou arxiv 2024 paper code data tl dr we introduce a comprehensive medical benchmark to evaluate openai o1 s clinical capabilities and report preliminary results indicating stronger clinical reasoning together with known limitations and actionable directions for improvement brain tumor segmentation through supervoxel transformer yunfei xie ce zhou jieru mei xianhang li cihang xie yuyin zhou isbi 2024 paper tl dr we developed two cnn transformer hybrid models as part of the brats isbi 2024 challenge to create brain tumor segmentation models with broad applicability education ph d in computer science rice university aug 2025 present houston tx b eng in artificial intelligence huazhong university of science technology sep 2021 jun 2025 wuhan china experience dec 2023 feb 2025 research intern vlaa university of california santa cruz supervisor prof yuyin zhou and prof cihang xie and dr jieru mei focus multimodal language models for medicine jul 2023 dec 2023 research intern ccvl johns hopkins university supervisor prof cihang xie and dr jieru mei focus multi level segmentation few shot medical segmentation reviewer 2026 iclr icml cvpr eccv 2025 iccv iclr cvpr icml tmm neurips tpami 2024 cvpr icml ieee isbi this website template was borrowed from jon barron last updated on march 16th 2026