Meta tags:
description=Ph.D. student at Peking University focusing on multimodal LLMs and OOD generalization.;

Headings (most frequently used words):

qixun, wang, 王启迅, news, selected, papers, see, full, publication, experience, education, awards,

Text of the page (most frequently used words):
and (27), wang (21), 2026 (12), qixun (11), visual (11), the (10), reasoning (9), peking (8), university (8), latent (8), pdf (7), for (6), code (6), technology (5), propose (5), arxiv (5), yisen (5), mllms (5), 2022 (4), learning (4), multimodal (4), language (4), zhang (4), models (4), training (4), ood (4), generalization (4), neurips (4), yifei (4), top (3), student (3), 2023 (3), 2025 (3), intelligence (3), science (3), vision (3), research (3), benchmark (3), generated (3), semantic (3), icml (3), out (3), distribution (3), theoretical (3), xianghua (3), ying (3), including (3), cvpr (3), 王启迅 (3), award (2), academic (2), merit (2), scholarship (2), outstanding (2), third (2), awards (2), computer (2), school (2), present (2), education (2), intern (2), march (2), kling (2), team (2), kuaishou (2), experience (2), framework (2), ignorance (2), prior (2), both (2), preprint (2), evaluating (2), artifacts (2), videos (2), yang (2), shi (2), artifact (2), bench (2), assessing (2), reveal (2), jing (2), enriched (2), adversarial (2), with (2), failure (2), invariant (2), graph (2), data (2), new (2), significant (2), algorithm (2), tasks (2), analysis (2), iclr (2), space (2), papers (2), publication (2), accepted (2), paper (2), was (2), may (2), excellence, 2021, innovation, yanchuang, capital, graduate, class, eecs, 2019, candidate, machine, working, agents, 快手科技, dual, perspective, diagnostic, vlms, revealing, how, routing, failures, undermine, model, grounding, validity, runyu, zhou, diagnosing, comprehensive, ability, detect, analyze, yuqi, tang, zhuoran, group, researchers, detecting, address, lack, richness, embeddings, learned, paradigms, tianrun, yue, sun, jingyi, yuan, tianren, longteng, guo, fengyun, rao, lyu, feng, chen, liu, simple, yet, effective, low, rank, strategy, that, improves, recognition, hong, zhu, improving, structured, priors, spotlight, theoretically, empirically, demonstrate, modes, classic, approaches, objective, performance, gains, guarantees, 2024, dissecting, graphs, capability, boundary, selection, mechanism, icl, through, carefully, designed, experiments, can, context, really, generalize, 200, github, stars, dataset, construction, sft, algorithms, achieving, improvements, domain, benchmarks, yuanxing, pengfei, wan, kun, gai, monet, beyond, images, selected, see, full, two, were, covering, unified, february, joined, our, one, released, video, detection, news, design, across, behavior, llms, agentic, evaluation, large, interests, include, year, blog, about, toggle, navigation,

Text of the page (random words):
qixun wang 王启迅 qixun wang 王启迅 toggle navigation about publication experience education awards blog qixun wang 王启迅 i am a third year ph d student in the school of intelligence science and technology at peking university my research interests include multimodal large language models mllms including latent visual reasoning agentic visual reasoning and evaluation of mllms out of distribution ood generalization including theoretical analysis and algorithm design across computer vision graph data and the generalization behavior of llms news may 2026 one paper artifact bench was released on arxiv for assessing mllms in ai generated video detection may 2026 our paper semantic enriched latent visual reasoning was accepted to icml 2026 march 2026 i joined the kling team at kuaishou technology as a research intern february 2026 two papers were accepted to cvpr 2026 covering reasoning in latent visual space and a benchmark for unified multimodal models selected papers see full publication cvpr monet reasoning in latent visual space beyond images and language qixun wang yang shi yifei wang yuanxing zhang pengfei wan kun gai xianghua ying and yisen wang cvpr 2026 propose a new framework for multimodal latent reasoning including dataset construction sft and rl algorithms achieving significant improvements on both in domain and ood visual reasoning benchmarks 200 github stars pdf code iclr can in context learning really generalize to out of distribution tasks qixun wang yifei wang xianghua ying and yisen wang iclr 2025 reveal the capability boundary and algorithm selection mechanism of icl on ood tasks through carefully designed experiments and theoretical analysis pdf code neurips dissecting the failure of invariant learning on graphs qixun wang yifei wang yisen wang and xianghua ying neurips 2024 theoretically and empirically demonstrate the failure modes of classic invariant learning approaches on graph data and propose a new training objective with significant performance gains and theoretical guarantees pdf code neurips spotlight improving out of distribution generalization by adversarial training with structured priors qixun wang yifei wang hong zhu and yisen wang neurips 2022 propose a simple yet effective low rank adversarial training strategy that improves the ood generalization of visual recognition models pdf code icml semantic enriched latent visual reasoning tianrun xu yue sun qixun wang jingyi lu yuan wang tianren zhang longteng guo fengyun rao jing lyu feng chen and jing liu icml 2026 reveal and address the lack of semantic richness in latent embeddings learned by prior latent visual reasoning training paradigms pdf code arxiv artifact bench evaluating mllms on detecting and assessing the artifacts of ai generated videos yuqi tang yang shi zhuoran zhang qixun wang and a group of outstanding researchers arxiv preprint 2026 propose a comprehensive benchmark for evaluating mllms ability to detect and analyze artifacts in ai generated videos pdf code arxiv diagnosing visual ignorance in vision language models runyu zhou qi zhang qixun wang and yisen wang arxiv preprint 2026 propose a dual perspective diagnostic framework for visual ignorance in vlms revealing how language prior routing failures undermine both model grounding and benchmark validity pdf experience kling team kuaishou technology 快手科技 research intern march 2026 present working on multimodal agents education ph d candidate in machine learning and computer vision school of intelligence science and technology peking university 2023 present b s in intelligence science and technology eecs peking university 2019 2023 awards the third class scholarship of peking university 2025 merit student at peking university 2025 outstanding graduate of peking university 2023 yanchuang capital scholarship top 6 2022 merit student at peking university top 6 2022 academic innovation award at peking university top 1 2022 award for academic excellence 2021 2026 qixun wang