EMNLP 2026 main</p>\n","updatedAt":"2026-08-24T05:29:07.003Z","author":{"_id":"66decf61f9971122eec44dc8","avatarUrl":"/avatars/ffd1bf114f2fffc1f9a2ffe5964543f3.svg","fullname":"Enjun Du","name":"EnjunDu","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":3,"isUserFollowing":false}},"numEdits":1,"identifiedLanguage":{"language":"en","probability":0.7493515610694885},"editors":["EnjunDu"],"editorAvatarUrls":["/avatars/ffd1bf114f2fffc1f9a2ffe5964543f3.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.20886","authors":[{"_id":"6a8bd5fe3d26296ea30919eb","name":"Enjun Du","hidden":false},{"_id":"6a8bd5fe3d26296ea30919ec","name":"Siyi Liu","hidden":false},{"_id":"6a8bd5fe3d26296ea30919ed","name":"Zirong Chen","hidden":false},{"_id":"6a8bd5fe3d26296ea30919ee","name":"Xinyu Zuo","hidden":false},{"_id":"6a8bd5fe3d26296ea30919ef","name":"Jinwen Luo","hidden":false},{"_id":"6a8bd5fe3d26296ea30919f0","name":"Ruiwen Tao","hidden":false},{"_id":"6a8bd5fe3d26296ea30919f1","name":"Lisheng Duan","hidden":false},{"_id":"6a8bd5fe3d26296ea30919f2","name":"Haijin Liang","hidden":false},{"_id":"6a8bd5fe3d26296ea30919f3","name":"Jin Ma","hidden":false},{"_id":"6a8bd5fe3d26296ea30919f4","name":"Junfu Pu","hidden":false},{"_id":"6a8bd5fe3d26296ea30919f5","name":"Yongqi Zhang","hidden":false}],"publishedAt":"2026-08-21T00:00:00.000Z","submittedOnDailyAt":"2026-08-24T00:00:00.000Z","title":"EviRank: Structured Relevance Evidence for Multimodal Image Re-ranking","submittedOnDailyBy":{"_id":"66decf61f9971122eec44dc8","avatarUrl":"/avatars/ffd1bf114f2fffc1f9a2ffe5964543f3.svg","isPro":false,"fullname":"Enjun Du","user":"EnjunDu","type":"user","name":"EnjunDu"},"summary":"Real-world image search queries are multimodal and compositional: ``find this shirt in pink'' specifies an entity to retain, an attribute to modify, and context to ignore. Yet existing re-rankers either compress such multifaceted relevance into an opaque embedding or rely on free-form chain-of-thought that easily omits or hallucinates fine-grained constraints. Drawing on rubric- and checklist-based evaluation from NLP, we recast multimodal image re-ranking as a semantic constraint satisfaction problem and propose EviRank, which parses any query - text-only, image-only, or composed - into a unified evidence package: typed criteria across six semantic slots (e.g., entities, attributes, relations), each labelled required, forbidden, or ignorable. Re-ranking then reduces to evidence-conditioned verification, combining deterministic rubric scoring and evidence-grounded listwise comparison in a single training-free procedure. The explicit evidence can further serve as structured supervision for optionally distilling a lightweight student. Across five benchmarks spanning text-to-image, image-to-image, and composed image retrieval, EviRank achieves state-of-the-art performance, and the distilled student preserves over 90% of the teacher's capability at substantially lower cost.","upvotes":8,"discussionId":"6a8bd5fe3d26296ea30919f6","ai_summary":"EviRank reformulates multimodal image re-ranking as semantic constraint satisfaction by parsing queries into structured evidence packages and verifying candidates via rubric scoring and listwise comparison without training.","ai_keywords":["multimodal image re-ranking","compositional queries","semantic constraint satisfaction","evidence package","rubric scoring","listwise comparison","training-free","distilled student"],"ai_summary_model":"thinkingmachines/Inkling-Small"},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"66decf61f9971122eec44dc8","avatarUrl":"/avatars/ffd1bf114f2fffc1f9a2ffe5964543f3.svg","isPro":false,"fullname":"Enjun Du","user":"EnjunDu","type":"user"},{"_id":"66d05337e62d6bbf50186c2f","avatarUrl":"/avatars/f5be15e754f0fbbb37d2cc5ea417f729.svg","isPro":false,"fullname":"Yijie Jin","user":"DrewJin0827","type":"user"},{"_id":"668fcd97495674543600db0b","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/JfBeJ6gSXe379G9ejjrLR.jpeg","isPro":false,"fullname":"Gao Mingqi","user":"EchoMinkki","type":"user"},{"_id":"65df4f5af3efe60b06ea1409","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65df4f5af3efe60b06ea1409/u-9mnMU8bNXvyQ8z1CG5m.png","isPro":false,"fullname":"Xinye Li","user":"asdfo123","type":"user"},{"_id":"687f853bb39262ba84f3eeff","avatarUrl":"/avatars/cdfc44fde8237f08f10192553fe5a075.svg","isPro":false,"fullname":"Junhao Shen","user":"shenjunhao","type":"user"},{"_id":"67d62da597767f49259a55d6","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/3bsqzf_J0Kv9tZ3qvZsE8.png","isPro":false,"fullname":"Xiangnan Wu","user":"XiangnanW","type":"user"},{"_id":"66699aa8a33847217b5a49c7","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/u8Z-6U8U7ARXOpdBDI7Qm.png","isPro":false,"fullname":"Weijie Wang","user":"lhmd","type":"user"},{"_id":"6971ec48ba9e9917b8861c15","avatarUrl":"/avatars/cc68fd9e0d4b696cbedea41b96f71ddf.svg","isPro":false,"fullname":"Haomin Wan","user":"Nek0laos","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.20886.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2608.20886 in a model README.md to link it from this page.

Cite arxiv.org/abs/2608.20886 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2608.20886 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.