[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-frameforge-adaptive-frame-gap-image-generation":3,"topics-all":38,"news-related-bf2c4f0f-3911-4227-b741-7252b1875cc7":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},"bf2c4f0f-3911-4227-b741-7252b1875cc7","FrameForge:视频模型跨行图生图,帧距自己挑","FrameForge把统一图生图塞进冻结的视频大模型:时序专家按输入挑虚拟帧距,推理潜变量补上条件到目标的变换,一个LoRA一并通吃主体生成、边缘图生图与指令编辑。论文自评编辑总分 6.12 反超任务专属基线,动作类编辑 6.84 超过表中闭源的 Gemini 2.5,登上 HuggingFace 当日论文榜榜首。","统一图生图一直有个工程默认值:主体生成、边缘图生图、指令编辑,一类任务养一个适配器。想用一个模型通吃,视频底座看起来是天然候选——它天生就在学「帧与帧之间怎么变」。10 月 27 日挂上 arXiv 的 FrameForge([arXiv:2610.36545](https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.36545))是这个思路的新版本:冻结 HunyuanVideo1.5,只训一个 LoRA,通吃三类任务。一作 Yijing Lin 也是前作 RealGeneral 的一作。这篇论文昨日才由作者本人提交到 Hugging Face 社区榜,现在已经排在当日论文榜榜首,拿到 89 个赞。\n\n## 固定帧距是个被忽略的假设\n\n把图生图塞进视频模型的已有做法(RealGeneral、DRA-Ctrl、ChronoEdit 这些)大多把条件帧和目标帧钉在固定的时序位置上,所有样本共用同一个「帧距」。论文的观察是:这个默认值站不住。编辑任务通常只改一小部分;主体生成和边缘图生图却可能大改外观、结构、场景。论文用 LPIPS 分布展示:三类任务「条件-目标」的变化幅度明显不同,而且单个任务内部的分布也很宽——任务标签根本说不清每个样本该变多少。他们的解法:帧距不写死,让模型按输入自己挑。\n\n## MoTE:三个时序专家,路由器挑帧距\n\n具体实现叫 Mixture of Temporal Experts。三个时序 RoPE 专家对应三档虚拟帧距,等效视频帧跨度 {17, 69, 121};一个轻量路由器读文本、条件和带噪目标的表示,样本级 Top-1 选一个专家,直通估计器让离散路由也能端到端训练,再配一条负载均衡损失防止路由塌缩。最有意思的是路由行为分析:论文对 500 个样本、50 个去噪步做了统计,canny 这类变化幅度大的任务在去噪早期偏向长距专家;主体生成和编辑先选中距,随着目标内容浮现转向短距,编辑转得更早更彻底。也就是说,路由不是给任务贴标签,而是随去噪进程动态变。\n\n## CoTR:一个没人监督的「中间帧」\n\n另一个模块解决「怎么变」。CoTR 在条件和目标之间插一个噪声初始化的推理潜变量,让它与文本、条件、目标互相做注意力,逐步聚上下文;中间训练阶段用一个随机注意力掩码(概率 0.5)切断目标对文本\u002F条件的直接注意,逼信息走这条潜变量通路。全程没有中间帧监督,只靠目标帧的扩散损失带着学。消融还显示潜变量不是越多越好:两个、三个反而全面劣化,一个就够。\n\n## 账面数字:自评里赢在哪,输在哪\n\n训练成本不算夸张:底座全冻结,秩 32 的 LoRA,可训练参数 182M,三任务各 170K 样本共 510K,2 张 A800 训了约 6 天。以下结果全部是论文自评。DreamBench 主体生成总分 0.739,在对比方法里最高,比第二的 ChronoEdit(0.729)高 1.37%,但主体一致性 0.739 仍排在 ChronoEdit(0.757)之后。COCO2017 边缘图生图 SSIM 0.447 是表里最高,比第二名 EasyControl 高 7.2%;FID 23.47 却不敌 ControlNet 的 20.42,论文自己也承认 FID 与 SSIM 度量的东西不同,得合着看。编辑是亮点:论文表格里 GEdit-Bench 开源阵营总分 6.12 最高,动作类编辑 6.84,表格中这一项超过了闭源的 Gemini 2.5(6.72)和 Seedream 4.0(5.53),仍低于 GPT-4o 的 7.81;总分与闭源三家的 7.81-8.01 也还差一档。\n\n## 消融裂缝:单补一件不够,合用才反超\n\n消融表里最有信息量的一行是:共享 LoRA 的统一基线把编辑从任务专属的 5.99 拖到 4.94——统一化确实在打架。单加 MoTE 回到 5.77,单加 CoTR 回到 5.56,谁都补不回 5.99;两个合用 6.12,反超任务专属。「选帧距」和「怎么变」是两件互补的事,只修一件都不够。换成 HunyuanVideo1.0 底座方向一致(编辑 4.38 到 5.00),说明思路不绑定单一底座。代码和 LoRA 权重已放出,许可证沿用腾讯混元社区许可。\n\n所以这篇论文真正值得记的不是「视频模型又赢了一次」,而是它把「条件到目标该变多少」从人工先验变成了模型自己路由的输入相关量。论文对比表里,除同样用视频底座的 ChronoEdit 外,其余开源方法动作类编辑得分全部落在 1.13 到 3.12——图像原生那一批缺的可能恰恰是「变化幅度」的常识。至于 6.12 对 8.01 的差距什么时候补上,那是下一个适配器要回答的问题。","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.36545","7437aeb9-930c-4866-a2e9-48003c1a792b",[11,15,18,21],{"id":12,"name":13,"slug":13,"description":14,"color":14},"7b67033c-19e6-4052-a626-e681bba64c7a","diffusion",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"d11f0044-8aef-487c-bebe-89ce4683a4a3","moe",{"id":19,"name":20,"slug":20,"description":14,"color":14},"c883fd20-1d66-4fb7-9fc7-320fa7f87023","text-to-image",{"id":22,"name":23,"slug":23,"description":14,"color":14},"ebe5dcd1-46b1-4298-b8c2-8e0e2f456e56","video-generation",[25],{"id":26,"lang":27,"title":28,"summary":29,"content":30},"99c169bd-364d-4c54-9841-2c90db05e797","en","FrameForge: Video Models Pick Their Own Frame Gaps","One frozen video model, one LoRA: FrameForge unifies three image tasks. In its table, editing 6.12 beats task-specific; motion 6.84 tops Gemini 2.5.","Unified image generation has a long-standing engineering default: subject-driven generation, canny-to-image, and instruction editing each keep their own adapter. If one model is to cover all of them, a video backbone looks like a natural host — it already learns how frames change into each other. FrameForge ([arXiv:2610.36545](https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.36545)), posted on arXiv on Oct 27, is the newest take on this idea: freeze HunyuanVideo1.5, train a single LoRA, and cover all three tasks. Yijing Lin, the lead author, also headed RealGeneral. The authors submitted the paper to the Hugging Face community board only yesterday; it now sits at the top of the daily papers board with 89 upvotes.\n\n## The fixed frame gap is an overlooked assumption\n\nExisting crossovers (RealGeneral, DRA-Ctrl, ChronoEdit) mostly pin the condition frame and the target frame to fixed temporal positions, so every sample shares one \"frame gap\". The paper's observation: this default does not hold. Editing usually changes a small part; subject-driven generation and canny-to-image can overhaul appearance, structure, or the whole scene. The paper's LPIPS distributions show the condition-target variation magnitudes differ clearly across tasks and stay wide within each task — a task label cannot say how much an individual sample should change. Their answer: do not hardcode the gap; let the model pick per input.\n\n## MoTE: three temporal experts, a router picks the gap\n\nThe implementation is a Mixture of Temporal Experts. Three temporal RoPE experts cover three virtual frame gaps with equivalent video-frame spans of 17, 69, and 121. A lightweight router reads the text, condition, and noisy-target representations and picks one expert per sample (top-1); a straight-through estimator keeps discrete routing trainable end to end, and a load-balancing loss prevents collapse. The routing analysis is the most interesting part: over 500 cases and 50 denoising steps, canny — a task with large visual changes — favors the long-interval expert early; subject-driven generation and editing start on the medium interval and shift to the short one as target content emerges, with editing switching earlier and harder. Routing does not tag tasks; it moves with the denoising process.\n\n## CoTR: one unsupervised \"middle frame\"\n\nThe other module handles \"how to change\". CoTR inserts a noise-initialized reasoning latent between the condition and the target, letting it attend to text, condition, and target to gather context step by step. During the intermediate training stage, a stochastic attention mask (probability 0.5) cuts the target's direct attention to text and condition, forcing information through this latent pathway. There is no intermediate-frame supervision anywhere — only the target-frame diffusion loss drives the learning. The ablation also shows more latents are not better: two or three degrade all three tasks, and one is enough.\n\n## The numbers: where it wins and where it loses (all paper-reported)\n\nThe training bill is moderate: backbone fully frozen, a rank-32 LoRA, 182M trainable parameters, 170K samples per task (510K total), 2 A800 GPUs for about 6 days. DreamBench subject-driven overall 0.739, the highest among the compared methods, 1.37% over ChronoEdit (0.729), while subject consistency 0.739 still trails ChronoEdit (0.757). On COCO2017 canny-to-image, SSIM 0.447 is the best in the table, 7.2% over EasyControl; FID 23.47 loses to ControlNet's 20.42 — the authors themselves note FID and SSIM measure different things and should be read together. Editing is the highlight: in the paper's table on GEdit-Bench, the open-source best overall is 6.12, and motion-change editing reaches 6.84, above closed-source Gemini 2.5 (6.72) and Seedream 4.0 (5.53), still below GPT-4o (7.81); overall it also trails the three closed-source systems (7.81-8.01).\n\n## The ablation crack: one fix is not enough, two together overtake\n\nThe most informative row of the ablation table: a shared-LoRA unified baseline drags editing from the task-specific 5.99 down to 4.94 — unification genuinely interferes. Adding MoTE alone recovers to 5.77, CoTR alone to 5.56; neither reaches 5.99. Together they hit 6.12 and overtake the task-specific score. \"Picking the gap\" and \"how to change\" are complementary jobs; fixing only one is not enough. Switching to a HunyuanVideo1.0 backbone keeps the direction (editing 4.38 to 5.00), so the idea does not depend on one backbone. Code and LoRA weights are released, under the Tencent Hunyuan Community License.\n\nWhat is worth remembering is not \"video models won again\", but that \"how much the condition should change\" went from a hand-set prior to an input-dependent, routed quantity. In the paper's comparison table, every open-source method except ChronoEdit — also video-backed — scores between 1.13 and 3.12 on motion editing; what image-native stacks lack may be exactly this sense of change magnitude. When the 6.12-to-8.01 gap closes is a question for the next adapter.","frameforge-adaptive-frame-gap-image-generation","2026-11-02T17:13:17Z","2026-11-02T17:13:25.235269Z","2026-11-02T17:13:25.235276Z",true,"agent",283,[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},"65fc5af8-80a4-4ba0-b9b9-6714d0968c25","第三方压H3到13.7B:消费卡9.1GB权重跑768p","hyperflow-h3-flash-int8-consumer-gpu-video","2026-11-06T08:30:00+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"06b504ab-f9ac-46b2-a1e0-acfbe0fd3996","上交开源WaveForcing:流式生视频134.8帧\u002F秒","wave-forcing-streaming-video-generation","2026-11-03T13:05:45+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"85b85139-0d39-464a-99cf-415fd498e9d2","L3G-World:高斯记忆让世界模型9 FPS实时生长","l3g-world-gaussian-memory-realtime","2026-10-30T21:12:08+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"703ab8f9-4f6c-46fd-8b3c-f53387619cad","更稀疏、更清晰:视频扩散注意力砍掉九成","ab-attention-block-sparse-video-diffusion","2026-10-30T15:15:10+00:00",{"id":80,"title":81,"news_slug":82,"published_at":83},"20f285d4-4a2f-4722-973a-465e3da4c5e1","蚂蚁IDM蒸馏:2步出图,长文本渲染94.25%","idm-intrinsic-distribution-matching-longtext","2026-10-30T11:01:03+00:00",{"id":85,"title":86,"news_slug":87,"published_at":88},"b860aced-d186-4d56-b751-53d5ca1ed658","OmniShowcase:4张参考图,主体转360度不变形","omnishowcase-360-subject-video","2026-10-29T23:06:48+00:00"]