[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-worldcast-benchmark-world-state-prediction":3,"topics-all":38,"news-related-7010d17a-2089-48fe-88b5-cd1eaef0de72":57},{"id":4,"title":5,"summary":6,"content":7,"original_url":8,"source_id":9,"tags":10,"translations":24,"news_slug":31,"published_at":32,"created_at":33,"modified_at":34,"is_published":35,"publish_type":36,"image_url":14,"view_count":37},"7010d17a-2089-48fe-88b5-cd1eaef0de72","WorldCast基准:GPT-Image-2.5拿72.8分,最强开源只有20.1","美团实习期间提出 WorldCast 基准:320 案例专考图像视频模型多步世界状态预测,12 模型评测,GPT-Image-2.5 得 72.8,Nano Banana Pro 42.8,最强开源 SenseNova-U1 仅 20.1;Agent 循环把最高分推到 82.4,瓶颈正转向语言模型的规划推理。","让图像模型「画未来」,比让它画得好看难得多:给人几张现状图和一本说明书,要求它按顺序画出接下来每一步的世界状态,组装步骤、物理变化、规则推演都得对。美团实习期间完成的研究提出 WorldCast 基准,专考这种多步世界状态预测——论文把模型能力分为 L1-L4 四级,指出高层任务探索仍然不足,WorldCast 瞄准 L3-L4。[arXiv:2610.20731]\n\n## 320 个案例,三个维度打分\n\n基准包含 320 个人工设计案例、8 个任务类别,合计要求生成 1,352 个视觉状态。类别不按「画什么」分,而按「决定未来过程的信息编码在哪里、需要什么推理才能转成状态转移」来划分,覆盖物理模拟、自适应任务完成、视觉目标实现、上下文程序学习等。一个 L4 案例要求模型读懂家具组装说明书,自己规划组装顺序,渲染出每个中间状态,并保持工作台面和零件外观一致——项目页展示的结果里,GPT-Image-2.5 完成了关键里程碑,Wan 3.0 无法遵循组装顺序。\n\n评分用三个互补维度:任务完成度(TC)、因果合理性(CP,转移是否遵守物理法则和逻辑规则)、视觉一致性(VC,身份、数量、几何、材质是否保持)。GPT-5.5 任裁判按 0-3 分打分,最终分为三项加权重定标。\n\n## 榜单:断层比想象的大\n\n12 个图像和视频生成系统同协议评测。GPT-Image-2.5 System(配 GPT-6 Astra)拿 72.8 分,论文称其在真实任务上展现早期视觉状态预测能力;Nano Banana Pro 只有 42.8;最强开源模型 SenseNova-U1 总分 20.1——头部闭源与最强开源之间隔着超过 3.5 倍的落差。\n\n难度爬升时分得更开:GPT-Image-2.5 全部五个 L4 类别都超过 60 分,L4 保留 71.6,只比 L3 低 2.2;SenseNova-U1 却从 L3 的 24.4 掉到 L4 的 15.0。级别越高,差距越大。失败模式具体:意外旋转、错误的镜面反射、密度错误、形状漂移、组装顺序错乱。\n\n## 三个发现\n\n**生成模型的强,很大程度是语言底座的强。** 把 Nano Banana Pro 配上 GPT-6 Astra 做增强,总分从 42.8 涨到 61.2,仍落后 GPT-Image 家族;把 GPT-Image-2 的底座从 GPT-5.5 换成 GPT-6 只有边际收益,物理模拟一项还退步。生成能力的上限被语言模型的规划推理焊住。\n\n**输出格式是隐形变量。** GPT-Image-2.5 从交错式输出切换成网格输出,总分掉 22.7,逻辑与规则类从 74.3 跌到 23.0。四步生成中,弱模型随上下文累积退化,GPT-Image-2.5 保持稳定。\n\n**Agent 化还能再抬天花板。** 研究者用 Codex 搭了两个等预算 harness:检查-修订循环把 GPT-Image-2.5 从 72.8 推到 80.2,把 GPT-6 增强版 Nano Banana Pro 从 61.2 推到 72.5;代码优先循环先构造并检查中间视觉工件再生成,把最高总分推到 82.4——代码把中间状态显式化了。\n\n## 所以呢\n\n这份基准的真正信号不是「GPT-Image-2.5 又赢了」,而是图像生成竞赛的重心正从「画得像」移到「推得对」,后者恰好是语言模型的主场。开源侧 20.1 的总分说明高层世界建模任务上开源还没真正入场,要补的可能不是扩散主干,而是规划和推理层。数据集在 Hugging Face(wenhaoyan77\u002FWorldCast),项目页 teal024.github.io\u002FWorldCast 有完整榜单。\n\n当「画得好看」已经饱和,下一个分水岭是:你的模型知不知道下一步世界该长什么样?","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.20731","7437aeb9-930c-4866-a2e9-48003c1a792b",[11,15,18,21],{"id":12,"name":13,"slug":13,"description":14,"color":14},"120fa59a-ff6f-4537-9bf5-f818df636a0e","benchmark",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"499f4b56-819d-49a3-9609-33e775143b86","multimodal",{"id":19,"name":20,"slug":20,"description":14,"color":14},"42e59a88-7795-47dc-a334-ef1e72c24347","openai",{"id":22,"name":23,"slug":23,"description":14,"color":14},"c883fd20-1d66-4fb7-9fc7-320fa7f87023","text-to-image",[25],{"id":26,"lang":27,"title":28,"summary":29,"content":30},"47b5fdde-18e6-4856-bb0e-6a57e5d925a8","en","WorldCast: GPT-Image-2.5 Scores 72.8, Best Open Model 20.1","WorldCast tests multi-step world state prediction on 320 cases: GPT-Image-2.5 scores 72.8, Nano Banana Pro 42.8, and the best open model only 20.1.","Asking an image model to \"draw the future\" is far harder than asking it to draw well. Given a few observation images and a manual, the model must render each successive world state in order — assembly steps, physical changes, and rule inference all have to be right. This task, multi-step world state prediction, demands tightly entangled understanding and generation, plus the world knowledge to translate human action sequences, physical transformations, and abstract visual reasoning into coherent frames. A study completed during an internship at Meituan proposes the WorldCast benchmark, organizes model capabilities into a four-level hierarchy (L1–L4), and argues that higher-level tasks remain underexplored despite rapid model progress. WorldCast targets exactly that L3–L4 band. [arXiv:2610.20731]\n\n## 320 Cases, Three Scoring Dimensions\n\nWorldCast contains 320 manually designed cases across eight task categories, requesting 1,352 visual states in total. Categories are not split by \"what to draw\" but by where the information governing a future process is encoded and what reasoning is needed to turn it into state transitions: physical simulation, adaptive task completion, visual goal realization, in-context procedural learning, and more. One L4 case asks the model to read a furniture assembly manual, plan the order itself, and render every intermediate state while preserving the work surface and component appearance. On the project page's shown result, GPT-Image-2.5 completes the key milestones while Wan 3.0 fails to follow the assembly order.\n\nScoring uses three complementary dimensions: Task Completion (TC, whether the sequence visibly realizes the task), Causal Plausibility (CP, whether transitions follow physical laws and logical rules), and Visual Consistency (VC, whether identity, count, geometry, and material persist). GPT-5.5 acts as the judge on a 0–3 scale; the final score is a weighted, rescaled combination.\n\n## The Leaderboard: A Bigger Cliff Than Expected\n\nTwelve image and video generation systems were evaluated under the same protocol. GPT-Image-2.5 System (paired with GPT-6 Astra) scores 72.8, which the paper describes as early-stage visual state prediction capability on real-world tasks. Nano Banana Pro reaches only 42.8. The best open-source model, SenseNova-U1, scores 20.1 — a gap of more than 3.5x between the top closed system and the best open weights.\n\nThe divergence grows with difficulty: GPT-Image-2.5 scores above 60 on all five L4 categories and retains 71.6 on L4, only 2.2 below its L3 score. SenseNova-U1 drops from 24.4 at L3 to 15.0 at L4. The higher the level, the wider the gap. Failure modes are concrete: unintended rotation, wrong reflections, erroneous density, shape drift, and incorrect assembly order.\n\n## Three Findings\n\n**Strong generators draw on their language backbones.** Pairing Nano Banana Pro with GPT-6 Astra as the enhancer raises its overall score from 42.8 to 61.2, yet it still trails the GPT-Image family. Conversely, swapping GPT-Image-2's backbone from GPT-5.5 to GPT-6 yields only marginal gains, with a decline in Physical Simulation. The ceiling of generative capability is welded shut by the planning and reasoning capacity of the language model.\n\n**Output format is a hidden variable.** Switching GPT-Image-2.5 from interleaved to grid output cuts its overall score by 22.7, with Logic and Rules falling from 74.3 to 23.0. Across four generation steps, weaker models degrade as context accumulates, while GPT-Image-2.5 stays stable.\n\n**Agentic harnesses raise the ceiling further.** The researchers built two Codex-based harnesses with the same generation budget: an inspect-and-revise loop without coding tools lifts GPT-Image-2.5 from 72.8 to 80.2 and GPT-6-enhanced Nano Banana Pro from 61.2 to 72.5. A coding-first loop that constructs and inspects intermediate visual artifacts before generation pushes the best overall score to 82.4, with the largest gains on Visual Goal Realization, Logic and Rules, and Physical Simulation — code makes intermediate states explicit.\n\n## So What\n\nThe real signal of this benchmark is not \"GPT-Image-2.5 won again,\" but that the image-generation race is shifting from \"drawing it well\" to \"reasoning it right\" — and the latter is the language model's home turf. The open-source side's 20.1 overall and 15.0 at L4 suggest open weights have not truly entered the high-level world-modeling game; catching up may require not a bigger diffusion backbone but better planning and reasoning layers. The dataset is on Hugging Face (wenhaoyan77\u002FWorldCast), and the project page (teal024.github.io\u002FWorldCast) carries the full leaderboard and case gallery.\n\nWhen \"looks good\" is saturated, the next dividing line is: does your model know what the world should look like next?","worldcast-benchmark-world-state-prediction","2026-10-19T15:09:58Z","2026-10-19T15:10:02.044926Z","2026-10-19T15:10:02.044937Z",true,"agent",76,[39,48],{"slug":40,"tag_slug":40,"title_zh":41,"title_en":42,"intro_zh":43,"intro_en":44,"id":45,"is_active":35,"created_at":46,"modified_at":47},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":49,"tag_slug":49,"title_zh":50,"title_en":51,"intro_zh":52,"intro_en":53,"id":54,"is_active":35,"created_at":55,"modified_at":56},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":58},[59,64,69,74,79,84],{"id":60,"title":61,"news_slug":62,"published_at":63},"63299a5b-f3b7-40c4-9985-0132e0cbffac","多轮图像编辑漂移的解法：MT Forcing 用自回归扩散把单轮质量带进多轮","mt-forcing-multi-turn-image-editing","2026-10-20T05:06:17+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"5757433d-e8ca-4371-89bc-f563dacc8089","程序能跑≠画对了:MaLiang框架专治AI作画跑偏","maliang-harness-p2v-gap","2026-09-30T15:20:00+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"aa9b279e-07c6-4d0c-86f0-403c4af321fa","蚂蚁 RULER:六维评分表给 SVG 生成重写奖励信号","ruler-rubric-rewards-svg-generation","2026-09-23T23:06:24+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"42d37ae0-3268-4684-91b1-9fca91f4e9c1","OpenAI 发布 ChatGPT Images 2.5:画个涂鸦就能出图,生成延迟砍半","openai-chatgpt-images-2-5-sketch","2026-09-09T19:30:00+00:00",{"id":80,"title":81,"news_slug":82,"published_at":83},"cf7aeaaa-7179-4de2-b177-9adf54668d61","StreamEval：37 模型实测，实时视频最缺听指令","streameval-interactive-streaming-video-benchmark","2026-10-22T11:08:09+00:00",{"id":85,"title":86,"news_slug":87,"published_at":88},"bbbda7ff-b1c7-4142-b37f-2b3fd0547e02","Fibo 2.0 开源:生成编辑合一,登顶开源编辑榜","bria-fibo-2-0-unified-editing-generation","2026-10-21T21:08:11+00:00"]