[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-helios-pku-bytedance-14b-19-5fps-long-video":3,"topics-all":36,"news-related-a16f5374-a870-42fd-8b5c-4e703a6ff31a":55},{"id":4,"title":5,"summary":6,"content":6,"original_url":7,"source_id":8,"tags":9,"translations":23,"news_slug":29,"published_at":30,"created_at":31,"modified_at":32,"is_published":33,"publish_type":34,"image_url":13,"view_count":35},"a16f5374-a870-42fd-8b5c-4e703a6ff31a","北京大学与字节跳动联合发布Helios：首个单卡19.5FPS实时生成长视频的14B模型","视频生成模型正进入实时长视频的新战场。北京大学、字节跳动、Canva等联合发布了Helios，这是一个140亿参数的统一视频生成模型，能够在单张H100 GPU上以19.5 FPS的速度实时生成长达分钟级的视频，且无需KV-cache、稀疏注意力或量化等标准加速技术。\n\n**核心突破在哪里？**\n\n传统视频生成模型在长视频场景下会面临漂移（drifting）问题——生成到中后期时画面质量严重下降。业界通常通过自强制（self-forcing）、误差库（error-banks）或关键帧采样来缓解，但这些方法引入额外计算开销。Helios的解法是从训练端入手：团队分析了典型的失败模式，在训练中显式模拟漂移场景，让模型从源头学会消除重复运动，而不是事后补救。\n\n另一个关键创新是效率。与小模型不同，14B参数模型此前无法在单卡上实时运行。Helios通过重压缩历史和噪声上下文，减少采样步数，使计算成本降至与13亿参数模型相当甚至更低，同时基础设施层面的优化进一步降低了显存占用——在80GB显存内可容纳4个14B模型做并行训练。\n\n**实际性能如何？**\n\n在短视频（~121帧）和长视频（~1440帧）两个维度上，Helios均超越此前的蒸馏模型，且推理速度大幅领先同规模方案。它支持T2V（文生视频）、I2V（图生视频）和V2V（视频生视频）三种任务，统一建模在单一架构内。\n\n团队宣布将开源代码、基础模型和蒸馏模型。这意味着社区可以在消费级硬件上复现和优化这一能力，对视频生成技术的普及意义深远。\n\nHelios的出现验证了一个重要趋势：大模型推理并非只能用小模型+蒸馏的妥协路径。通过训练策略与系统优化的联合设计，14B级别的模型同样可以在成本可控的情况下实现实时交互。这对多模态Agent、视频编辑工具乃至实时视频对话场景，都具有结构性影响。原文：https:\u002F\u002Farxiv.org\u002Fabs\u002F2603.04379 | https:\u002F\u002Fpku-yuangroup.github.io\u002FHelios-Page","https:\u002F\u002Farxiv.org\u002Fabs\u002F2603.04379","7437aeb9-930c-4866-a2e9-48003c1a792b",[10,14,17,20],{"id":11,"name":12,"slug":12,"description":13,"color":13},"0ef8513a-0a26-42f0-b6f9-5b6dadded45c","efficiency",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"0a93ec8e-ea39-4693-81de-563ca8c173f7","inference",{"id":18,"name":19,"slug":19,"description":13,"color":13},"01598627-1ea6-4b27-a5d8-874971571a71","llm",{"id":21,"name":22,"slug":22,"description":13,"color":13},"ebe5dcd1-46b1-4298-b8c2-8e0e2f456e56","video-generation",[24],{"id":25,"lang":26,"title":27,"summary":28,"content":13},"6d2ecef4-da8b-48f0-a3c1-8be3ad110999","en","Helios: PKU and ByteDance's 14B real-time long-video model","Video generation models are entering a new battlefield: real-time long video. Peking University, ByteDance, Canva, and others have jointly released Helios, a 14-billion-parameter unified video generation model capable of real-time, minute-long video generation at 19.5 FPS on a single H100 GPU, without requiring standard acceleration techniques like KV-cache, sparse attention, or quantization.\n\n**Where is the core breakthrough?**\n\nTraditional video generation models face a drifting problem in long-video scenarios — by mid-to-late generation, image quality seriously degrades. The industry usually mitigates this through self-forcing, error-banks, or key-frame sampling, but these methods introduce extra compute overhead. Helios' solution is from the training side: the team analyzed typical failure modes, explicitly simulating drifting scenarios in training, letting the model learn to eliminate repetitive motion from the source, rather than post-hoc remedies.\n\nAnother key innovation is efficiency. Unlike small models, 14B-parameter models previously couldn't run real-time on a single card. Helios reduces compute cost to be on par with or lower than 1.3B-parameter models by recompressing history and noise context, reducing sampling steps, while infrastructure-level optimization further reduces VRAM usage — 4 14B models can fit in 80GB VRAM for parallel training.\n\n**How does actual performance stack up?**\n\nAcross both short video (~121 frames) and long video (~1440 frames) dimensions, Helios surpasses previous distilled models, with inference speed significantly leading same-scale solutions. It supports T2V (text-to-video), I2V (image-to-video), and V2V (video-to-video) tasks, unified-modeled within a single architecture.\n\nThe team announced it will open-source code, base models, and distilled models. This means the community can reproduce and optimize this capability on consumer-grade hardware, with profound implications for the popularization of video generation technology.\n\nHelios' emergence validates an important trend: large-model inference doesn't have to use only the compromise path of small model + distillation. Through joint design of training strategy and system optimization, 14B-class models can also achieve real-time interaction at controllable cost. This has structural impact on multimodal Agents, video editing tools, and even real-time video dialogue scenarios. Original: https:\u002F\u002Farxiv.org\u002Fabs\u002F2603.04379 | https:\u002F\u002Fpku-yuangroup.github.io\u002FHelios-Page","helios-pku-bytedance-14b-19-5fps-long-video","2026-04-29T08:05:00Z","2026-04-29T16:05:53.812761Z","2026-08-19T02:08:40.142862Z",true,"agent",178,[37,46],{"slug":38,"tag_slug":38,"title_zh":39,"title_en":40,"intro_zh":41,"intro_en":42,"id":43,"is_active":33,"created_at":44,"modified_at":45},"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":47,"tag_slug":47,"title_zh":48,"title_en":49,"intro_zh":50,"intro_en":51,"id":52,"is_active":33,"created_at":53,"modified_at":54},"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":56},[57,62,67,72,77,82],{"id":58,"title":59,"news_slug":60,"published_at":61},"c3814f7d-2649-4660-a798-28fb03aa2b6d","SwitchSD 让投机解码学会「该抄才抄」:读内部信号,EAGLE3 之上再快 15%","switchsd-copy-intent-speculative-decoding","2026-09-20T23:09:25+00:00",{"id":63,"title":64,"news_slug":65,"published_at":66},"3cecce90-70b9-4bb3-b9b7-93e6b0c05105","D-Quant 用熵编码压 KV:2.26bit 近无损","d-quant-entropy-coding-kv-cache","2026-09-20T17:10:42+00:00",{"id":68,"title":69,"news_slug":70,"published_at":71},"813ad679-51dd-43d7-afcc-0baf48d2ef5f","When2Think:推理模型该想多久,先看题有多难","when2think-difficulty-aware-length-control","2026-09-19T19:08:00+00:00",{"id":73,"title":74,"news_slug":75,"published_at":76},"ff6f65e1-28b2-4a48-b317-7870072ecfa9","VC-Attention低比特注意力:视频生成提速1.59倍","vc-attention-low-bit-video-attention","2026-09-17T13:30:00+00:00",{"id":78,"title":79,"news_slug":80,"published_at":81},"9a3cd449-e29a-4730-814b-f1be5c2685c6","复旦FFD让Flash Attention退役？11.6× kernel提速把长上下文推到256K","fudan-ffd-long-context-attention-sparsity","2026-09-15T07:15:46+00:00",{"id":83,"title":84,"news_slug":85,"published_at":86},"2638aeac-dc4d-4b73-b7fe-2b042015adee","OreoLook 开源:三层缓存把 AI 搜索搬进 8 核 CPU,重复问题 0.1 毫秒出答案","oreolook-three-layer-cpu-cache","2026-09-10T23:08:36+00:00"]