[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-mt-forcing-multi-turn-image-editing":3,"topics-all":38,"news-related-63299a5b-f3b7-40c4-9985-0132e0cbffac":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},"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","商汤等团队提出 MT Forcing 训练范式，用单轮分布匹配蒸馏加教师强制引导，压制自回归扩散在多轮图像编辑中的误差累积。在 MTEdit-Bench 上，八轮编辑的指令遵循从 8.47 提到 9.66，内容一致性从 7.09 提到 8.40，像素漂移从 6.34 降到 0.74。模型权重已开源。","做一张图很少只改一次。加个物件、调下光线、再修一处细节——真正难的是让前面已经满意的部分别被后面的编辑顺手毁掉。自回归扩散类模型在多轮编辑里最典型的翻车方式就是漂移：纹理一点点糊掉，主体一点点跑偏，几轮之后画面面目全非。\n\n一篇刚放出不久的论文给出了针对这个问题的训练范式。商汤研究团队联合港科大 MMLab、斯坦福与清华 AIR 的 MT Forcing，把训练放在模型自己生成的多轮轨迹上，同时用一条\"单轮教师线\"做监督——教师永远从原图出发，把至今累积的所有指令一次性应用完，再拿它的输出分布去校正学生。论文与项目页分别在 [arXiv:2610.25404](https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.25404) 与 [项目主页](https:\u002F\u002Fhouyuanchen111.github.io\u002Fmt-forcing.github.io\u002F) 上。\n\n## 漂移到底怎么发生的\n\n论文把多轮编辑的失败拆成两类：一类是视觉质量不稳定，伪影随轮次累积、细节退化；另一类是内容不一致，本该保留的区域和此前的编辑结果被意外改动。当模型持续编辑自己的输出时，微小误差会被一步步放大。\n\n这套训练范式基于统一多模态模型（UMM）上的自回归扩散，作者列出两个互补的训练信号，分别对应上面两类漂移。\n\n## 两条信号：一个稳住画质，一个守住一致性\n\n**单轮分布匹配蒸馏（Single-Turn DMD）** 盯的是画质。学生按自己的输出历史逐轮编辑自己，而教师从原始图像出发，把到当前为止的全部指令一次性合成一遍，给出分布层面的指导。这样教师不会把上一轮的误差带到下一轮，避免递归式误差累积。\n\n**教师强制引导（Teacher-Forcing Guidance）** 盯的是跨轮内容一致性。它在真实的多轮轨迹上监督依赖历史的转场：预热阶段用参考轨迹引导，蒸馏阶段作为正则项继续施加。作者明确指出，只靠逐轮的分布匹配不足以保证跨轮一致——这也是为什么需要第二条信号。\n\n两条信号作用在同一个学生模型上。\n\n## 数字：八轮编辑下的变化\n\n项目页披露的数据来自 MTEdit-Bench 的 100 组八轮编辑会话，取八轮平均分，基座是 SenseNova-U1.5：\n\n- **指令遵循（IF）**：基座 8.47 → 加上 MT Forcing 后 **9.66**（满分 10）\n- **内容一致性（CC）**：基座 7.09 → **8.40**\n- **抗漂移（DRS）**：基座 3.83 → **4.30**（满分 5）\n- **匹配点位移**：基座 6.34 px → **0.74 px**（越低越好）\n\n评估口径项目页也写清楚了：八轮平均等权计分，IF 与 CC 采用该基准\"失败即触发\"的规则——某轮失败之后的后续轮次直接记零分。\n\n第 8 轮的内容一致性差距更直观：7.68 对基座的 3.22，高出 4.46 分。\n\n## 我的判断\n\n这篇工作有意思的地方不在某个分数，而在于它指出了一个容易被忽略的训练信号设计问题：**多轮场景下，教师的定义方式本身会决定误差是否累积**。多数方法默认教师和学生处在同一轮次，于是教师也带着上一轮的误差往前爬；MT Forcing 把教师从原图重新出发、一次性合并全部指令，本质是让监督信号对历史误差免疫。\n\n两点得说清楚。第一，仓库代码尚未释出，官方标注是\"即将发布\"，训练代码与预训练 checkpoint 都还没落到仓库里，想立刻复现的人还得等。第二，实验结果来自团队自建基准 MTEdit-Bench 的自报数字，第三方复现尚未见到，这是评测类工作该打的折扣。\n\n多轮编辑正从\"能演示\"走向\"能交付\"——用户不会接受改了五次之后画面已经认不出来了。谁能把\"记住前面发生过什么\"训练进模型，谁就拿到交互式视觉创作的入口。","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.25404","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},"7b67033c-19e6-4052-a626-e681bba64c7a","diffusion",{"id":19,"name":20,"slug":20,"description":14,"color":14},"499f4b56-819d-49a3-9609-33e775143b86","multimodal",{"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},"b31ff89a-226d-48de-8031-4964e4ea1859","en","Stabilizing Multi-Turn Image Editing with MT Forcing","SenseTime-led MT Forcing curbs drift in autoregressive multi-turn image editing: eight-turn instruction following 8.47 to 9.66 on MTEdit-Bench.","Rarely does a finished image stop at a single edit. You add an object, adjust the lighting, then refine a detail — and the real difficulty is keeping the parts you were already happy with from being quietly destroyed by the next pass. Autoregressive diffusion models drift in exactly this way during multi-turn editing: textures blur little by little, the subject shifts slightly, and after a few turns the image is unrecognizable.\n\nA recently released paper proposes a training paradigm aimed squarely at that failure mode. MT Forcing, from SenseTime Research with HKUST MMLab, Stanford and Tsinghua AIR, trains on the model's own multi-turn rollouts while drawing supervision from a single-turn teacher line — the teacher always starts from the original image, applies every accumulated instruction in one pass, and its output distribution corrects the student. The paper and project page are available at [arXiv:2610.25404](https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.25404) and on the [project site](https:\u002F\u002Fhouyuanchen111.github.io\u002Fmt-forcing.github.io\u002F).\n\n## How the drift happens\n\nThe paper splits multi-turn editing failures into two categories: one is unstable visual quality, where artifacts accumulate across turns and details degrade; the other is inconsistent content, where regions that should be preserved and earlier edits get changed unintentionally. When a model continuously edits its own outputs, small errors compound step by step.\n\nThe framework builds on unified multimodal models (UMMs) with autoregressive diffusion. The authors describe two complementary training signals, one for each failure category.\n\n## Two signals: one holds quality, one holds consistency\n\n**Single-Turn Distribution Matching Distillation (Single-Turn DMD)** targets visual quality. The student edits its own generated history one turn at a time, while the teacher starts from the original image, composes all instructions accumulated so far in a single pass, and provides distribution-level guidance. That way the teacher carries no error from one turn into the next, avoiding recursive error accumulation. The project page summarizes it as a single-turn teacher present at every turn.\n\n**Teacher-Forcing Guidance** targets cross-turn content consistency. It supervises history-dependent transitions along real multi-turn trajectories: reference trajectories guide the warmup phase, and the signal continues as a regularizer during distillation. The authors state explicitly that per-turn distribution matching alone cannot guarantee cross-turn consistency — which is exactly why the second signal is needed.\n\nBoth signals act on the same student model, each covering one side of the problem.\n\n## The numbers across eight turns\n\nThe disclosed figures come from 100 eight-turn sessions in MTEdit-Bench, scored as eight-turn averages with SenseNova-U1.5 as the base model:\n\n- **Instruction following (IF)**: 8.47 base → **9.66** with MT Forcing (out of 10)\n- **Content consistency (CC)**: 7.09 → **8.40**\n- **Drift resistance (DRS)**: 3.83 → **4.30** (out of 5)\n- **Matched-point displacement**: 6.34 px → **0.74 px** (lower is better)\n\nThe project page also spells out the evaluation protocol: eight-turn means weight every turn equally, and IF and CC follow the benchmark's failure-triggered scoring rule, where turns after a failure score zero. Pixel metrics measure mean displacement between matched points, with lower values indicating better spatial alignment, and each metric is interpreted separately.\n\nLooking at turn 8 alone makes the gap starker: 7.68 against the base model's 3.22, a 4.46-point margin. The project page also provides a fifteen-turn editing comparison demo.\n\n## Reading it\n\nThe interesting part here is not any single score but a training-signal design problem that tends to get overlooked: in multi-turn settings, how the teacher is defined determines whether error compounds. Most methods assume teacher and student sit at the same turn, so the teacher climbs forward carrying its own accumulated error. MT Forcing instead sends the teacher back to the original image and folds all instructions into one pass, which in effect makes the supervision signal immune to historical error.\n\nTwo caveats are worth stating plainly. First, the repository has not released code yet — it is officially marked as coming soon, with training code, inference code and pretrained checkpoints all absent, so anyone wanting to reproduce immediately will have to wait. Second, the reported numbers are team-published results on MTEdit-Bench, the authors' own benchmark; third-party replication has not appeared so far, which is the discount one should apply when reading evaluation work of this kind.\n\nMulti-turn editing is moving from demonstrable to deliverable — users will not accept an image that no longer looks like itself after five edits. Whoever trains \"remembering what came before\" into the model holds the actual entry point for interactive visual creation.","mt-forcing-multi-turn-image-editing","2026-10-20T05:06:17Z","2026-10-20T05:06:21.878777Z","2026-10-20T05:06:21.878785Z",true,"agent",86,[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},"7010d17a-2089-48fe-88b5-cd1eaef0de72","WorldCast基准:GPT-Image-2.5拿72.8分,最强开源只有20.1","worldcast-benchmark-world-state-prediction","2026-10-19T15:09:58+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"3bf416bb-a989-4b02-8341-dbafc7c96c48","MOSS-Image 开源:5B 双语生图,文字排版可编程","moss-image-bilingual-text-rendering-base","2026-10-18T19:08:44+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"473ca44d-b179-4c19-a7a6-916dd8b90e4a","FLUX 3 Image 发布:画框控图,逐框编辑","flux-3-image-bounding-boxes","2026-10-02T23:06:31+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"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":80,"title":81,"news_slug":82,"published_at":83},"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":85,"title":86,"news_slug":87,"published_at":88},"5ff06769-2251-40fd-a672-f394a1f68965","十人合影谁是谁:腾讯混元 WithEveryone 给群像生成装上身份锚点","witheveryone-group-image-identity-grounding","2026-08-24T13:00:00+00:00"]