[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-pcs-llm-progressive-transfer":3,"news-related-029d5b6c-a448-442b-b742-96afeaab330f":36},{"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},"029d5b6c-a448-442b-b742-96afeaab330f","PCS 把 LLM 推理能力\"渐进迁移\"到任意语种：5 个语种验证，轻量翻译替代昂贵蒸馏","NJU\u002FIAAR 团队提出 PCS（Progressive Code-Switching）框架：通过轻量翻译构造代码切换推理轨迹做 SFT 初始化，再用带 step-level 语言一致性课程学习的强化学习逐步提高目标语种占比，让 LRM 直接在 5 种类型学差异较大的语言上输出连贯的多步推理。整套流程不需要更强的 LRM 蒸馏数据，也无需外部 judge 模型在线评估，把多语推理迁移从\"高成本依赖\"压成\"轻量翻译+GRPO 风格 RL\"。在多语言推理基准上显著缩小目标语种与英语的差距，是英语 LRM 多语扩张的可复制路径。","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.00485","7437aeb9-930c-4866-a2e9-48003c1a792b",[10,14,17,20],{"id":11,"name":12,"slug":12,"description":13,"color":13},"5e628969-6d2a-437f-998a-104e4b16cfb1","ai-progress",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"40269b40-7942-4650-9672-ed2e6524d37a","ai-technology",{"id":18,"name":19,"slug":19,"description":13,"color":13},"0ef8513a-0a26-42f0-b6f9-5b6dadded45c","efficiency",{"id":21,"name":22,"slug":22,"description":13,"color":13},"01598627-1ea6-4b27-a5d8-874971571a71","llm",[24],{"id":25,"lang":26,"title":27,"summary":28,"content":13},"9eb3bb9a-3364-4f34-877d-b5c568d34e89","en","PCS progressively transfers reasoning to any language","The NJU\u002FIAAR team proposes the PCS (Progressive Code-Switching) framework: through lightweight translation constructing code-switched reasoning traces for SFT initialization, then using step-level language-consistency curriculum-learning reinforcement learning to gradually increase the target language proportion, letting LRMs directly output coherent multi-step reasoning in 5 typologically diverse languages. The entire process doesn't need stronger LRM distillation data, nor external judge models for online evaluation, compressing multilingual reasoning migration from \"high-cost dependency\" to \"lightweight translation + GRPO-style RL\". On multilingual reasoning benchmarks, it significantly narrows the gap between target language and English, and is a replicable path for English LRMs' multilingual expansion.","pcs-llm-progressive-transfer","2026-07-08T14:15:00Z","2026-07-08T14:20:18.784988Z","2026-08-19T02:08:40.142862Z",true,"agent",66,{"items":37},[38,43,48,53,58,63],{"id":39,"title":40,"news_slug":41,"published_at":42},"90fc8caa-c5f6-45ab-adb8-50f28f43739b","字节 UP：正向 advantage 不裁剪，GRPO\u002FDAPO\u002FGSPO 即插即用","bytedance-seed-up-advantage","2026-07-08T04:21:42+00:00",{"id":44,"title":45,"news_slug":46,"published_at":47},"e4c13922-29e1-41a9-9470-8dae80f62368","推理模型的「无效思考」:55% 的 CoT 步骤对答案概率毫无影响","epiphenomenal-cot-55pct-useless-thinking","2026-06-14T10:01:00+00:00",{"id":49,"title":50,"news_slug":51,"published_at":52},"58645289-9914-4c61-a9bd-3691afa52dff","QK-Restore：给混合注意力LLM装上\"长程记忆保险丝\"，CoT微调后256K检索从65.4%拉回76.4%","qk-restore-long-range-memory-fuse-256k-76pct","2026-06-10T08:20:00+00:00",{"id":54,"title":55,"news_slug":56,"published_at":57},"43eda321-b0b7-4df7-b20e-9758cbab42c9","记忆越完整,眼前题越做不对:MemTrapBench 把 LLM 长期记忆框架打回原形","memtrapbench-llm-memory-cognitive-traps","2026-08-22T04:00:00+00:00",{"id":59,"title":60,"news_slug":61,"published_at":62},"deac2d55-76a6-40d2-8ef7-36aed2ad0105","Linux 7.2 把 AI 拉进内核开发:Sashiko 让补丁数量翻倍,Torvalds 接受「新常态」","linux-7-2-sashiko-ai-kernel-review","2026-08-20T12:00:00+00:00",{"id":64,"title":65,"news_slug":66,"published_at":67},"22a1a718-0eb6-46e5-8ee8-825400de11d1","DeepMind WeatherNext 在 Nature 发论文：用 28 km 粗分辨率做出多一天的飓风预警,代码权重全部开源","deepmind-weathernext-cyclones-nature-open-source","2026-08-10T02:00:00+00:00"]