{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,6]],"date-time":"2026-05-06T06:18:37Z","timestamp":1778048317976,"version":"3.51.4"},"reference-count":129,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,3,6]],"date-time":"2026-03-06T00:00:00Z","timestamp":1772755200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,3,6]],"date-time":"2026-03-06T00:00:00Z","timestamp":1772755200000},"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":[[2026,3,6]]},"DOI":"10.1109\/wacv61042.2026.00144","type":"proceedings-article","created":{"date-parts":[[2026,5,5]],"date-time":"2026-05-05T19:59:32Z","timestamp":1778011172000},"page":"1415-1427","source":"Crossref","is-referenced-by-count":0,"title":["MageBench: Bridging Large Multimodal Models to Agents"],"prefix":"10.1109","author":[{"given":"Miaosen","family":"Zhang","sequence":"first","affiliation":[{"name":"Southeast University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Qi","family":"Dai","sequence":"additional","affiliation":[{"name":"Microsoft"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yifan","family":"Yang","sequence":"additional","affiliation":[{"name":"Microsoft"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jianmin","family":"Bao","sequence":"additional","affiliation":[{"name":"Microsoft"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Dongdong","family":"Chen","sequence":"additional","affiliation":[{"name":"Microsoft"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kai","family":"Qiu","sequence":"additional","affiliation":[{"name":"Microsoft"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Chong","family":"Luo","sequence":"additional","affiliation":[{"name":"Microsoft"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xin","family":"Geng","sequence":"additional","affiliation":[{"name":"Southeast University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Baining","family":"Guo","sequence":"additional","affiliation":[{"name":"Southeast University"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","year":"2024","journal-title":"Build with Claude"},{"key":"ref2","year":"2024","journal-title":"Hello GPT-4o"},{"key":"ref3","year":"2024","journal-title":"Grok-2 beta release"},{"key":"ref4","year":"2024","journal-title":"Introducing OpenAI o1-preview"},{"key":"ref5","article-title":"Phi-3 technical report: A highly capable language model locally on your phone","author":"Abdin","year":"2024"},{"key":"ref6","article-title":"Gpt-4 technical report","author":"Achiam","year":"2023"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3551349.3559555"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref9","article-title":"Qwen2.5-VL technical report","author":"Bai","year":"2025"},{"key":"ref10","article-title":"Deepseek llm: Scaling open-source language models with longtermism","author":"Bi","year":"2024"},{"key":"ref11","article-title":"Language models are few-shot learners","volume-title":"NeurIPS","author":"Brown"},{"key":"ref12","article-title":"Partnr: A benchmark for planning and reasoning in embodied multi-agent tasks","volume-title":"ICLR","author":"Chang"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/3664647.3680575"},{"key":"ref14","article-title":"Evaluating large language models trained on code","author":"Chen","year":"2021"},{"key":"ref15","article-title":"Can vlms play action role-playing games? take black myth wukong as a study case","author":"Chen","year":"2024"},{"key":"ref16","article-title":"Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks","volume-title":"CVPR","author":"Chen"},{"key":"ref17","article-title":"Palm: Scaling language modeling with pathways","author":"Chowdhery","year":"2023","journal-title":"JMLR"},{"key":"ref18","article-title":"Training verifiers to solve math word problems","author":"Cobbe","year":"2021"},{"key":"ref19","article-title":"Nvlm: Open frontier-class multimodal llms","author":"Dai","year":"2024"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.478"},{"key":"ref21","article-title":"A survey on incontext learning","volume-title":"EMNLP","author":"Dong"},{"key":"ref22","article-title":"The llama 3 herd of models","author":"Dubey","year":"2024"},{"key":"ref23","article-title":"How far are we from agi","author":"Feng","year":"2024"},{"key":"ref24","article-title":"Blink: Multimodal large language models can see but not perceive","author":"Fu","year":"2024"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1145\/3664647.3681249"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1093\/logcom\/5.6.763"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.670"},{"key":"ref28","article-title":"rStar-Math: Small llms can master math reasoning with self-evolved deep thinking","author":"Guan","year":"2025"},{"key":"ref29","article-title":"Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning","author":"Guo","year":"2025"},{"key":"ref30","article-title":"Large language model based multi-agents: A survey of progress and challenges","author":"Guo","year":"2024"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.211"},{"key":"ref32","article-title":"Measuring coding challenge competence with apps","author":"Hendrycks","year":"2021"},{"key":"ref33","article-title":"Measuring mathematical problem solving with the math dataset","volume-title":"NeurIPS","author":"Hendrycks"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.5040\/9781501365072.13409"},{"key":"ref35","article-title":"Open-Reasoner-Zero: An open source approach to scaling reinforcement learning on the base model","author":"Hu","year":"2025"},{"key":"ref36","article-title":"Look before you leap: Unveiling the power of gpt-4v in robotic vision-language planning","author":"Hu","year":"2023"},{"key":"ref37","article-title":"Inner monologue: Embodied reasoning through planning with language models","author":"Huang","year":"2022"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"ref39","article-title":"TableVQA-Bench: A visual question answering benchmark on multiple table domains","author":"Kim","year":"2024"},{"key":"ref40","article-title":"Large language models are zero-shot reasoners","volume-title":"NeurIPS","author":"Kojima"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i04.5878"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1145\/3573010"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01263"},{"key":"ref44","article-title":"Mmro: Are multimodal llms eligible as the brain for in-home robotics?","author":"Li","year":"2024"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1126\/science.abq1158"},{"key":"ref46","article-title":"Appagent v2: Advanced agent for flexible mobile interactions","author":"Li","year":"2024"},{"key":"ref47","author":"Lightman","year":"2023","journal-title":"Let\u2019s verify step by step"},{"key":"ref48","article-title":"Mcu: A task-centric framework for open-ended agent evaluation in minecraft","author":"Lin","year":"2023"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00566"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02484"},{"key":"ref51","author":"Liu","year":"2024","journal-title":"LLaVA-NeXT: Improved reasoning, ocr, and world knowledge"},{"key":"ref52","article-title":"Visual instruction tuning","volume-title":"NeurIPS","author":"Liu"},{"key":"ref53","article-title":"Agentbench: Evaluating llms as agents","author":"Liu","year":"2023"},{"key":"ref54","article-title":"Visualagentbench: Towards large multimodal models as visual foundation agents","author":"Liu","year":"2024"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72658-3_13"},{"key":"ref56","article-title":"Deepseek-vl: towards real-world vision-language understanding","author":"Lu","year":"2024"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.acl-long.528"},{"key":"ref58","article-title":"Learn to explain: Multimodal reasoning via thought chains for science question answering","volume-title":"NeurIPS","author":"Lu"},{"key":"ref59","article-title":"Mathvista: Evaluating mathematical reasoning of foundation models in visual contexts","author":"Lu","year":"2023"},{"key":"ref60","volume-title":"DeepScaleR: Surpassing o1-preview with a 1.5b model by scaling rl","author":"Luo","year":"2025"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/ICRA57147.2024.10610855"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.findings-acl.177"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01367"},{"key":"ref64","article-title":"Adaptive machine translation with large language models","author":"Moslem","year":"2023"},{"key":"ref65","article-title":"Long-horizon planning for multi-agent robots in partially observable environments","volume-title":"NeurIPS","author":"Nayak"},{"key":"ref66","article-title":"Mobileflow: A multimodal llm for mobile gui agent","author":"Nong","year":"2024"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0312"},{"key":"ref68","article-title":"Summarization is (almost) dead","author":"Pu","year":"2023"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00886"},{"key":"ref70","article-title":"Mutual reasoning makes smaller llms stronger problem-solvers","author":"Qi","year":"2024"},{"key":"ref71","article-title":"Imagination-augmented agents for deep reinforcement learning","volume-title":"NeurIPS","author":"Racani\u00e8re"},{"key":"ref72","article-title":"Scaling language models: Methods, analysis & insights from training gopher","author":"Rae","year":"2021"},{"key":"ref73","article-title":"Code generation with alphacodium: From prompt engineering to flow engineering","author":"Ridnik","year":"2024"},{"key":"ref74","article-title":"A systematic survey of prompt engineering in large language models: Techniques and applications","author":"Sahoo","year":"2024"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2022.118669"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00943"},{"key":"ref77","article-title":"Proximal policy optimization algorithms","author":"Schulman","year":"2017"},{"key":"ref78","article-title":"Deepseekmath: Pushing the limits of mathematical reasoning in open language models","author":"Shao","year":"2024"},{"key":"ref79","article-title":"Reflexion: Language agents with verbal reinforcement learning","volume-title":"NeurIPS","author":"Shinn"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01075"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00280"},{"key":"ref82","article-title":"Aligning large multimodal models with factually augmented rlhf","author":"Sun","year":"2023"},{"key":"ref83","article-title":"Gemini: a family of highly capable multimodal models","author":"Anil","year":"2023"},{"key":"ref84","article-title":"Llama: Open and efficient foundation language models","author":"Touvron","year":"2023"},{"key":"ref85","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.23919\/FUSION59988.2024.10706514"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.1016\/j.jai.2024.12.003"},{"key":"ref88","article-title":"Mobile-agent: Autonomous multi-modal mobile device agent with visual perception","author":"Wang","year":"2024"},{"key":"ref89","article-title":"Measuring multimodal mathematical reasoning with math-vision dataset","volume-title":"NeurIPS","author":"Wang"},{"key":"ref90","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.1036"},{"key":"ref91","article-title":"MobileAgentBench: An efficient and user-friendly benchmark for mobile llm agents","author":"Wang","year":"2024"},{"key":"ref92","doi-asserted-by":"publisher","DOI":"10.1007\/s11704-024-40231-1"},{"key":"ref93","article-title":"Qwen2-vl: Enhancing vision-language model\u2019s perception of the world at any resolution","author":"Wang","year":"2024"},{"key":"ref94","article-title":"Self-consistency improves chain of thought reasoning in language models","author":"Wang","year":"2022"},{"key":"ref95","article-title":"Chain-of-thought prompting elicits reasoning in large language models","volume-title":"NeurIPS","author":"Wei"},{"key":"ref96","article-title":"mChartQA: A universal benchmark for multimodal chart question answer based on vision-language alignment and reasoning","author":"Wei","year":"2024"},{"key":"ref97","doi-asserted-by":"publisher","DOI":"10.1017\/S0269888900008122"},{"key":"ref98","doi-asserted-by":"publisher","DOI":"10.1007\/s11704-024-40013-9"},{"key":"ref99","article-title":"Dettoolchain: A new prompting paradigm to unleash detection ability of mllm","author":"Wu","year":"2024"},{"key":"ref100","article-title":"The rise and potential of large language model based agents: A survey","author":"Xi","year":"2023"},{"key":"ref101","article-title":"Logic-RL: Unleashing llm reasoning with rule-based reinforcement learning","author":"Xie","year":"2025"},{"key":"ref102","article-title":"A paradigm shift in machine translation: Boosting translation performance of large language models","author":"Xu","year":"2023"},{"key":"ref103","author":"Yang","year":"2024","journal-title":"Exploring diverse in-context configurations for image captioning"},{"key":"ref104","article-title":"Webshop: Towards scalable real-world web interaction with grounded language agents","volume-title":"NeurIPS","author":"Yao"},{"key":"ref105","article-title":"React: Synergizing reasoning and acting in language models","author":"Yao","year":"2022"},{"key":"ref106","article-title":"Minicpm-v: A gpt-4v level mllm on your phone","author":"Yao","year":"2024"},{"key":"ref107","doi-asserted-by":"publisher","DOI":"10.1093\/nsr\/nwae403"},{"key":"ref108","article-title":"Yi: Open foundation models by 01. ai","author":"Young","year":"2024"},{"key":"ref109","article-title":"Mm-vet: Evaluating large multimodal models for integrated capabilities","author":"Yu","year":"2023"},{"key":"ref110","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00913"},{"key":"ref111","article-title":"STaR: Self-taught reasoner bootstrapping reasoning with reasoning","volume-title":"NeurIPS","author":"Zelikman"},{"key":"ref112","article-title":"Transporter networks: Rearranging the visual world for robotic manipulation","volume-title":"CoRL","author":"Zeng"},{"key":"ref113","article-title":"Large language models for robotics: A survey","author":"Zeng","year":"2023"},{"key":"ref114","author":"Zhang","year":"2023","journal-title":"Prompting large language model for machine translation: A case study"},{"key":"ref115","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-78456-9_15"},{"key":"ref116","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-acl.73"},{"key":"ref117","article-title":"Internlm-xcomposer-2.5: A versatile large vision language model supporting long-contextual input and output","author":"Zhang","year":"2024"},{"key":"ref118","article-title":"Opt: Open pre-trained transformer language models","author":"Zhang","year":"2022"},{"key":"ref119","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00632"},{"key":"ref120","article-title":"Automatic chain of thought prompting in large language models","author":"Zhang","year":"2022"},{"key":"ref121","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.734"},{"key":"ref122","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.findings-acl.547"},{"key":"ref123","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0228"},{"key":"ref124","article-title":"Language agent tree search unifies reasoning acting and planning in language models","author":"Zhou","year":"2023"},{"key":"ref125","article-title":"Image-of-thought prompting for visual reasoning refinement in multimodal large language models","author":"Zhou","year":"2024"},{"key":"ref126","article-title":"Webarena: A realistic web environment for building autonomous agents","author":"Zhou","year":"2023"},{"key":"ref127","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-acl.940"},{"key":"ref128","article-title":"Ghost in the minecraft: Generally capable agents for open-world environments via large language models with text-based knowledge and memory","author":"Zhu","year":"2023"},{"key":"ref129","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.naacl-long.566"}],"event":{"name":"2026 IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV)","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 (WACV)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11491838\/11491925\/11492221.pdf?arnumber=11492221","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,6]],"date-time":"2026-05-06T05:55:38Z","timestamp":1778046938000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11492221\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,6]]},"references-count":129,"URL":"https:\/\/doi.org\/10.1109\/wacv61042.2026.00144","relation":{},"subject":[],"published":{"date-parts":[[2026,3,6]]}}}