Interleaved image-text reports are an important format for presenting complex multimodal information, yet generating them in a trustworthy and well-grounded way remains challenging. </p>\n<p><a href=\"https://cdn-uploads.huggingface.co/production/uploads/6710ac3fb4ee4920580a5f0e/wRA4asRvqS4HS71Yrl8B0.png\" rel=\"nofollow\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/6710ac3fb4ee4920580a5f0e/wRA4asRvqS4HS71Yrl8B0.png\" alt=\"image\"></a></p>\n<p>In this work, we introduce <strong>Ptah</strong>, an agentic harness for producing reliable multimodal reports by coordinating textual research, claim-grounded evidence, and source-aligned visual evidence. </p>\n<p><a href=\"https://cdn-uploads.huggingface.co/production/uploads/6710ac3fb4ee4920580a5f0e/i4LM12H2EBVGnsqNCO2RE.png\" rel=\"nofollow\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/6710ac3fb4ee4920580a5f0e/i4LM12H2EBVGnsqNCO2RE.png\" alt=\"image\"></a></p>\n<p>Evaluating multimodal reports is also difficult, as factual grounding, citation fidelity, visual relevance, cross-modal consistency, and presentation quality all matter. To address this, we propose <strong>PtahEval</strong>, an evaluation protocol for assessing multimodal report quality at both the image-content and presentation levels.</p>\n<p><a href=\"https://cdn-uploads.huggingface.co/production/uploads/6710ac3fb4ee4920580a5f0e/a5DGUeKguJAVAWAQC1Av0.png\" rel=\"nofollow\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/6710ac3fb4ee4920580a5f0e/a5DGUeKguJAVAWAQC1Av0.png\" alt=\"image\"></a></p>\n","updatedAt":"2026-05-29T05:57:13.753Z","author":{"_id":"6710ac3fb4ee4920580a5f0e","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6710ac3fb4ee4920580a5f0e/OhQQFlZmkmLQpMYqKCGP6.jpeg","fullname":"Chenghao Zhang","name":"SnowNation","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":16,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7773786187171936},"editors":["SnowNation"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6710ac3fb4ee4920580a5f0e/OhQQFlZmkmLQpMYqKCGP6.jpeg"],"reactions":[],"isReport":false}},{"id":"6a1a4164de0290e64fb905dd","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false},"createdAt":"2026-05-30T01:46:12.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"This is an automated message from the [Librarian Bot](https://huggingface.co/librarian-bots). I found the following papers similar to this paper. \n\nThe following papers were recommended by the Semantic Scholar API \n\n* [Deep-Reporter: Deep Research for Grounded Multimodal Long-Form Generation](https://huggingface.co/papers/2604.10741) (2026)\n* [ViDR: Grounding Multimodal Deep Research Reports in Source Visual Evidence](https://huggingface.co/papers/2605.13034) (2026)\n* [CogGen: A Cognitively Inspired Recursive Framework for Deep Research Report Generation](https://huggingface.co/papers/2604.17072) (2026)\n* [DR$^{3}$-Eval: Towards Realistic and Reproducible Deep Research Evaluation](https://huggingface.co/papers/2604.14683) (2026)\n* [InterLV-Search: Benchmarking Interleaved Multimodal Agentic Search](https://huggingface.co/papers/2605.07510) (2026)\n* [MTA-Agent: An Open Recipe for Multimodal Deep Search Agents](https://huggingface.co/papers/2604.06376) (2026)\n* [Towards On-Policy Data Evolution for Visual-Native Multimodal Deep Search Agents](https://huggingface.co/papers/2605.10832) (2026)\n\n\n Please give a thumbs up to this comment if you found it helpful!\n\n If you want recommendations for any Paper on Hugging Face checkout [this](https://huggingface.co/spaces/librarian-bots/recommend_similar_papers) Space\n\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: `@librarian-bot recommend`","html":"<p>This is an automated message from the <a href=\"https://huggingface.co/librarian-bots\">Librarian Bot</a>. I found the following papers similar to this paper. </p>\n<p>The following papers were recommended by the Semantic Scholar API </p>\n<ul>\n<li><a href=\"https://huggingface.co/papers/2604.10741\">Deep-Reporter: Deep Research for Grounded Multimodal Long-Form Generation</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.13034\">ViDR: Grounding Multimodal Deep Research Reports in Source Visual Evidence</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.17072\">CogGen: A Cognitively Inspired Recursive Framework for Deep Research Report Generation</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.14683\">DR$^{3}$-Eval: Towards Realistic and Reproducible Deep Research Evaluation</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.07510\">InterLV-Search: Benchmarking Interleaved Multimodal Agentic Search</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.06376\">MTA-Agent: An Open Recipe for Multimodal Deep Search Agents</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.10832\">Towards On-Policy Data Evolution for Visual-Native Multimodal Deep Search Agents</a> (2026)</li>\n</ul>\n<p> Please give a thumbs up to this comment if you found it helpful!</p>\n<p> If you want recommendations for any Paper on Hugging Face checkout <a href=\"https://huggingface.co/spaces/librarian-bots/recommend_similar_papers\">this</a> Space</p>\n<p> You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: <code><span class=\"SVELTE_PARTIAL_HYDRATER contents\" data-target=\"UserMention\" data-props=\"{&quot;user&quot;:&quot;librarian-bot&quot;}\"><span class=\"inline-block\"><span class=\"contents\"><a href=\"/librarian-bot\">@<span class=\"underline\">librarian-bot</span></a></span> </span></span> recommend</code></p>\n","updatedAt":"2026-05-30T01:46:12.547Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.6823925971984863},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2605.29861","authors":[{"_id":"6a19288856b4bb14ec65d0cc","user":{"_id":"6710ac3fb4ee4920580a5f0e","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6710ac3fb4ee4920580a5f0e/OhQQFlZmkmLQpMYqKCGP6.jpeg","isPro":false,"fullname":"Chenghao Zhang","user":"SnowNation","type":"user","name":"SnowNation"},"name":"Chenghao Zhang","status":"claimed_verified","statusLastChangedAt":"2026-05-29T08:49:35.607Z","hidden":false},{"_id":"6a19288856b4bb14ec65d0cd","name":"Guanting Dong","hidden":false},{"_id":"6a19288856b4bb14ec65d0ce","name":"Yufan Liu","hidden":false},{"_id":"6a19288856b4bb14ec65d0cf","name":"Tong Zhao","hidden":false},{"_id":"6a19288856b4bb14ec65d0d0","name":"Zhicheng Dou","hidden":false}],"publishedAt":"2026-05-28T00:00:00.000Z","submittedOnDailyAt":"2026-05-29T00:00:00.000Z","title":"Towards Verifiable Multimodal Deep Research: A Multi-Agent Harness for Interleaved Report Generation","submittedOnDailyBy":{"_id":"6710ac3fb4ee4920580a5f0e","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6710ac3fb4ee4920580a5f0e/OhQQFlZmkmLQpMYqKCGP6.jpeg","isPro":false,"fullname":"Chenghao Zhang","user":"SnowNation","type":"user","name":"SnowNation"},"summary":"Large Language Models (LLMs) have advanced autonomous agents from deep search, which retrieves concise factual answers, to deep research, which synthesizes scattered evidence into long-form reports. However, verifiable multimodal deep research remains challenging due to open-ended synthesis without deterministic ground truth and the need to interleave textual arguments with visual evidence. We propose Ptah, a multi-agent harness for interleaved report generation. Ptah orchestrates the lifecycle from user query to rendered web report through planning, research, and writing stages, where specialized agents construct visual-aware plans, collect claim-grounded evidence, maintain source-aligned images in a Visual Working Memory, and compose reports through declarative multimodal tool use. A verifier agent serves as the harness's acceptance function, enforcing factual grounding, citation fidelity, and cross-modal consistency throughout the workflow. We further introduce PtahEval, an evaluation protocol that augments existing benchmarks with image-level and presentation-level assessments. Experiments on deep research benchmarks show that Ptah produces more reliable, visually informative, and usable human-facing multimodal reports than strong baselines.","upvotes":7,"discussionId":"6a19288856b4bb14ec65d0d1","ai_summary":"Multi-agent system for generating reliable, visually informative multimodal reports by interleaving textual and visual evidence through specialized agents and verification mechanisms.","ai_keywords":["large language models","autonomous agents","deep search","deep research","multimodal deep research","multi-agent harness","visual working memory","declarative multimodal tool use","verifier agent","evaluation protocol","deep research benchmarks"],"organization":{"_id":"622177ac43826d6f261f8208","name":"RUC","fullname":"Renmin University of China","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/61ac8f8a00d01045fca0ad2f/670IAX9A2-BflqA5MiSBW.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6710ac3fb4ee4920580a5f0e","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6710ac3fb4ee4920580a5f0e/OhQQFlZmkmLQpMYqKCGP6.jpeg","isPro":false,"fullname":"Chenghao Zhang","user":"SnowNation","type":"user"},{"_id":"664c4ddf4bea570e25cb4cc9","avatarUrl":"/avatars/13c805437efd34c5e6b7a3a9c229696a.svg","isPro":false,"fullname":"Vincent zhao","user":"Tung111","type":"user"},{"_id":"64d068a231c655ff8a77153e","avatarUrl":"/avatars/2b7407be92b65d435fecc3c29e7f8455.svg","isPro":false,"fullname":"wenhan liu","user":"liuwenhan","type":"user"},{"_id":"61cd4b833dd34ba1985e0753","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/61cd4b833dd34ba1985e0753/BfHfrwotoMESpXZOHiIe4.png","isPro":false,"fullname":"KABI","user":"dongguanting","type":"user"},{"_id":"6764f88d80ac6b6be311cb0a","avatarUrl":"/avatars/e898cd1483d481df6dcb12751ab403f9.svg","isPro":false,"fullname":"Fiona Ray","user":"Fionawww","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"},{"_id":"6270324ebecab9e2dcf245de","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6270324ebecab9e2dcf245de/cMbtWSasyNlYc9hvsEEzt.jpeg","isPro":false,"fullname":"Kye Gomez","user":"kye","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"622177ac43826d6f261f8208","name":"RUC","fullname":"Renmin University of China","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/61ac8f8a00d01045fca0ad2f/670IAX9A2-BflqA5MiSBW.jpeg"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2605/2605.29861.md"}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2605.29861 in a model README.md to link it from this page.

Cite arxiv.org/abs/2605.29861 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2605.29861 in a Space README.md to link it from this page.

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