[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-queen-chess-language-model-self-distillation":3,"topics-all":38,"news-related-34d309c3-5757-469a-85ff-b524c8865037":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},"34d309c3-5757-469a-85ff-b524c8865037","Queen 模型边下棋边解说:七轮自蒸馏涨 900 Elo","普林斯顿发布 4B 棋类语言模型 Queen:把 LC0 引擎的沉默专家表征经交叉注意力注入指令模型,再用自然语言版 Bellman 更新做迭代自蒸馏,七轮 Elo 从 1782 涨到 2697。论文自报棋力与解题精度超过所有前沿大模型,参数量少三个数量级,解释连贯度接近 GPT-5.6-Sol。代码与模型已开源。","国际象棋引擎早就下得比人类好,但 Stockfish、LC0 这类引擎是\"沉默的专家\"——它告诉你最佳走法,不告诉你为什么。语言模型正好相反:解释起来头头是道,棋力却撑不起它的解说。普林斯顿团队最近发布的 Queen,试图把这两半拼起来:一个 4B 参数的棋类语言模型,一边以典型特级大师水准行棋,一边解释自己的走法和计划。\n\n## 给\"沉默专家\"装上嘴\n\nQueen 采用编码器-解码器架构,核心思路是把国际象棋界的\"沉默专家\"——LC0 引擎——作为编码器,通过交叉注意力接入一个指令微调语言模型。训练用了一套问答课程(question-answering curriculum),让语言模型学会从编码器的表征里\"读出\"棋类概念。这相当于让会说话的模型旁听一位大师的内心独白,再用自己的话复述出来。\n\n## 自然语言版的 Bellman 更新\n\n真正让棋力起飞的是迭代自蒸馏。团队设计了一个强化学习中 Bellman 更新的自然语言类比:模型先分析自己候选走法落子后的局面,把这些分析归纳成对当前局面的解释,再把解释蒸馏回模型自身。七轮迭代下来,Elo 从 1782 涨到 2697,提升超过 900 分。按论文自报,它的棋力和解题精度超过了所有前沿大模型,而参数量比对手少三个数量级;在基于 LM 的评测里,它的解释连贯度接近 GPT-5.6-Sol(high 档)。注意这些对比数字都来自作者自己的评测,尚无第三方复现。\n\n## 开源了,但还毛糙\n\n代码以 Apache 2.0 协议开源于 GitHub,模型放在 Hugging Face 的 princeton-nlp 集合里。不过 README 自己承认\"扩展文档和清理后的代码即将发布\",仓库目前只有 5 个 commit,GPU 推理依赖的 vLLM 组合\"仍在验证中\"——想立刻上手复现的人要有心理准备。\n\n## 所以呢\n\n这篇论文真正值得记的不是下棋,是一个配方:凡是存在\"沉默专家编码器\"的领域——游戏、机器人、电脑操作——都可以用\"专家表征 + 交叉注意力 + 迭代自蒸馏\"给语言模型装上领域推理能力。4B 模型在垂直领域打赢参数量大三个数量级的通用模型,这条路如果跑通,很多场景就不必干等更大的通用模型了。\n\n参考:arXiv:2610.03695(https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.03695);代码仓库 github.com\u002Fqueen-project\u002Fqueen","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.03695","7437aeb9-930c-4866-a2e9-48003c1a792b",[11,15,18,21],{"id":12,"name":13,"slug":13,"description":14,"color":14},"5e628969-6d2a-437f-998a-104e4b16cfb1","ai-progress",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"0ef8513a-0a26-42f0-b6f9-5b6dadded45c","efficiency",{"id":19,"name":20,"slug":20,"description":14,"color":14},"01598627-1ea6-4b27-a5d8-874971571a71","llm",{"id":22,"name":23,"slug":23,"description":14,"color":14},"b9bd9039-fcdb-41a8-b85b-fc1587def2b9","open-source",[25],{"id":26,"lang":27,"title":28,"summary":29,"content":30},"f37ff69c-9606-4b93-8c43-4071b5147364","en","Queen: 4B chess LM plays at GM level while explaining its moves","Princeton's 4B chess-language model Queen plays at Grandmaster level while explaining its moves, gaining 900 Elo over seven self-distillation rounds.","Chess engines have long outplayed humans, but engines like Stockfish and LC0 are silent experts: they tell you the best move, not why it is best. Language models are the opposite — fluent explanations backed by weak play. Queen, a 4B-parameter chess-language model from the Princeton NLP team, tries to glue the two halves together: it plays at the level of a typical Grandmaster while explaining its moves and plans.\n\n## Giving the Silent Expert a Voice\n\nQueen uses an encoder-decoder architecture. The core idea is to take chess's silent expert — the LC0 engine — as an encoder and plug it into an instruction-tuned language model through cross-attention. Training follows a question-answering curriculum that teaches the LM to read chess concepts out of the encoder's representations. Think of it as letting a talkative model sit in on a master's inner monologue, then retell it in its own words.\n\n## A Natural-Language Bellman Update\n\nWhat really sends the playing strength up is iterative self-distillation. The team built a natural-language analog of the Bellman update from reinforcement learning: the model analyzes the positions after its top candidate moves, consolidates those analyses into an explanation of the current position, and distills that explanation back into itself. Over seven iterations, its Elo climbed from 1782 to 2697 — a gain of more than 900 points. Per the authors' own evaluation, it surpasses all frontier models on both playing strength and puzzle accuracy with three orders of magnitude fewer parameters, and LM-based evaluations put its explanation coherence close to GPT-5.6-Sol (high). Note these comparison numbers all come from the authors' own benchmarks; no third-party replication exists yet.\n\n## Open-Sourced, Still Rough\n\nThe code is on GitHub under Apache 2.0, with models on Hugging Face under the princeton-nlp collection. But the README itself admits the \"extended README and clean code release\" is still coming: the repo has just 5 commits, and the vLLM setup required for GPU inference is \"still being validated.\" Anyone hoping to replicate immediately should calibrate expectations.\n\n## So What\n\nThe thing worth remembering from this paper is not chess — it is a recipe. Anywhere a silent expert encoder exists — games, robotics, computer use — you can bolt domain-specific reasoning onto a language model via expert representations, cross-attention, and iterative self-distillation. If a 4B model beating generalist models three orders of magnitude larger keeps working in vertical domains, many use cases no longer need to wait for the next, bigger general model.\n\nReference: arXiv:2610.03695 (https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.03695); code at github.com\u002Fqueen-project\u002Fqueen","queen-chess-language-model-self-distillation","2026-10-05T23:04:34Z","2026-10-05T23:06:20.244569Z","2026-10-05T23:06:20.244579Z",true,"agent",958,[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},"c361090b-291c-4357-9bd3-54845025f29d","CaRE-KD 让蒸馏学会看「教师脸色」:NeurIPS 2026 选中 IIT Delhi","care-kd-confidence-distillation-neurips-2609-36734","2026-10-05T03:00:00+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"2dbc7c0f-083a-47d4-ba9a-8a7d66b22ae0","亚马逊八阶段配方:后训练让 GLM-4.5-Air 反超官方版","amazon-rufus-air-post-training-recipe","2026-09-25T23:12:24+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"054e060c-e182-42a9-b3ed-229feb8ac0ac","2026 年的蒸馏长什么样:Hugging Face 拆解前沿模型三大范式","2026-distillation-three-paradigms","2026-07-09T12:00:00+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"ab2d6e9e-8890-4ae7-b6ca-8febc831a279","HPLT MultiSynt\u002FMT：4.8 万亿 token 多语种数据集","hplt-multisynt-multilingual-dataset","2026-07-07T16:01:00+00:00",{"id":80,"title":81,"news_slug":82,"published_at":83},"93dfc6f4-e4a9-47a3-aa66-b4b9cee864e7","Beyond LoRA 不只是口号：HF 给 40+ PEFT 方法拍下公平基准，OFT 在图像任务上反超 LoRA","beyond-lora-hf-peft-benchmark-oft","2026-06-18T12:00:00+00:00",{"id":85,"title":86,"news_slug":87,"published_at":88},"95e663a4-2cdf-454c-9787-b154fbd41909","TokenRouter:token 级路由提速 64 倍","tokenrouter-token-level-llm-routing","2026-10-09T21:11:28+00:00"]