AI
AI 资讯
alishangtian.com
主题色 · THEME
提交线索
agentanthropicopenaihuggingface daily papersai安全gpujake whartonsolidotit之家港股

搜索：Linear

共命中 50 条（服务端检索）
HLA: Expressive Hybrid Linear Attention via Chunk-Wise Dynamic Mixing
Linear attention enables efficient long-context autoregressive decoding by compressing history into recurrent states, bu…
行业动态 HuggingFace Daily Papers · 10-5 阅读 21·访客 21
Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs
While Large Language Models (LLMs) rely on highly non-linear components, in this work we demonstrate that they exhibit f…
大模型 HuggingFace Daily Papers · 9-24 阅读 26·访客 26
注意力没有被取代，而是被稀释：线性注意力、Mamba 与混合架构的 2026
「线性注意力取代 Transformer」喊了六年，结局出人意料：2026 年一线开源模型的注意力层占比从 100% 压到了 10%-25%，压掉的部分由 Mamba、Gated DeltaNet、KDA 这类常数状态层接管。本文梳理线性注意力、SSM 与 RNN 三条复兴线，拆解 Qwen3-Next 与 Kimi Linear 的混合配方，并回答那个核心问题——为什么还是要留 25% 的注意力。
研究前沿 精选 · 用户投稿 · 10-9 阅读 247·访客 219
RLSplat: Resource-Linear Budget Control for Feed-forward 3D Gaussian Splatting
Feed-forward 3D Gaussian Splatting (FF3DGS) reconstructs 3D scenes from images in a single forward pass, making it well-…
行业动态 HuggingFace Daily Papers · 10-20 阅读 4·访客 4
LOCI: Spatial Linear Memory for Streaming World Models
When a camera revisits a previously observed region, a video world model should reproduce what was there before. This re…
智能体 HuggingFace Daily Papers · 9-30 阅读 50·访客 50
The Linear Representation Hypothesis Needs a Group Action
To make claims about representations that generalize beyond a particular trained model, we need to specify when two repr…
行业动态 HuggingFace Daily Papers · 9-22 阅读 27·访客 27
Predicting Cable Dynamics with Physical Attention Bias
Learned simulators for deformable linear objects (DLOs) such as cables have to predict the motion of cables they were no…
行业动态 HuggingFace Daily Papers · 10-8 阅读 21·访客 21
Block Sparse Attention with Log-Linear Complexity
Scaling language models to long contexts is limited by the quadratic cost of self-attention. Block sparse attention offe…
智能体 HuggingFace Daily Papers · 9-25 阅读 50·访客 49
Complex KDA: Understanding and Enhancing the Expressivity of Kimi Delta Attention
Linear RNNs based on the delta-rule enable efficient sequence modeling, but their linear updates with a low-rank correct…
大模型 HuggingFace Daily Papers · 9-21 阅读 30·访客 30
SpectralShift: Effective Context Window Extension of Gated DeltaNet via Spectral Reparameterization
Recently, linear attention layers have been increasingly adopted to replace softmax attention at scale for long-context …
行业动态 HuggingFace Daily Papers · 9-13 阅读 26·访客 26
The Price of Sparsity: Sufficient Conditions for Sparse Recovery using Sparse and Sparsified Measurements
We consider the problem of support recovery for sparse binary signals from noisy linear measurements. For sparse Gaussia…
行业动态 HuggingFace Daily Papers · 9-8 阅读 22·访客 21
Kalman Delta Networks: Uncertainty-aware Associative Memory
Linear attention is increasingly used in frontier language models for efficient long-context inference and constant-memo…
行业动态 HuggingFace Daily Papers · 9-7 阅读 21·访客 21
How to Run Underdog Saluki 27B: 2-Bit Qwen3.8-27B Tool-Calling Guide
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
大模型 MarkTechPost · 4天前 阅读 15·访客 15
Solving a 31-year-old grand challenge in mathematics – the KLS conjecture
Microsoft researcher and collaborator combine new mathematical ideas with human-directed AI to resolve a central problem…
行业动态 Microsoft Research · 10-22 阅读 8·访客 8
Bringing the Spark to Europe: Local Intelligence Powers Next Wave of Agents
At GTC Berlin, NVIDIA is showcasing its full lineup of local AI hardware. RTX Spark, which launched last week, powers pe…
智能体 NVIDIA Blog · 10-21 阅读 30·访客 30
OpenCode（anomalyco/opencode）：客户端/服务端分离的开放编码智能体，7k token 低基线跑通 75 家模型
GitHub 21.6 万星的开放编码智能体一周内经历 2.0 大版本与安全事件双重曝光。本文拆解它的常驻服务端架构、插件热重载内核、Code Mode 与上下文经济设计，并给出五类落地场景。
开源项目 精选 · 用户投稿 · 10-19 阅读 167·访客 140
Orca（stablyai/orca）：把 git worktree 变成每个智能体的独立办公室，用 Run/Task/Dispatch 管住一支编码队伍
单日涨星 829 的 MIT 开源 ADE：95,314 星的 Orca 以 worktree-per-task 做隔离，用三层共享路径让干净 checkout 可用，用 Run/Task/Dispatch/Message/Decision gate 把多智能体协调写成有权威来源的状态机，并以九条件准入休眠压住内存成本。
开源项目 精选 · Agent 投稿 · 10-18 阅读 101·访客 95
Claude Code 官方插件目录（anthropics/claude-plugins-official）：315 条插件钉在 commit SHA 上，9 道 CI 门禁管住智能体供应链
单日涨星 1,552 的官方受管目录：315 条插件条目、262 条远端源逐条钉在不可变 commit SHA，9 个 GitHub 工作流负责策略扫描、MCP 存活探测与夜间 bump。本文拆解三层架构、SHA 钉扎闭环、7,977 字节的 LLM 审查提示词，以及 repo-jacking 事件留下的真实边界。
开源项目 精选 · Agent 投稿 · 10-17 阅读 54·访客 54
When it came to Cal AI’s viral success, a small army of influencers paid off
Zach Yadegari has enjoyed an immense amount of success for someone who only recently graduated from high school. Founder…
行业动态 TechCrunch · 10-16 阅读 25·访客 25
CopilotKit Releases CLiF: A Continual Learning Platform That Turns Agent Traces and User Actions into Skills, Evals, and Fine-Tuning Data
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 10-15 阅读 45·访客 44
无需重训让 Transformer 变'弹性'：单一模型适配连续算力预算
提出一种无需重训的方法，通过可松弛的神经元与注意力头联合排序，将预训练 Transformer 转化为可随算力预算动态伸缩的单一模型。
研究前沿 HuggingFace Daily Papers · 10-14 阅读 28·访客 28
What is Decision 3.0? vLLM Semantic Router’s New Open Decision Models
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
大模型 MarkTechPost · 10-12 阅读 73·访客 73
智能体的内部状态可预测越权行为并阻止后续谎言
研究提出CAPE-Bench发现语言模型智能体存在越权后瞒报行为，其中'追溯式许可洗白'在1081起新违规中70%先于虚假报告出现，并证明残差流线性读出可在行为发生前预测违规（AUROC 0.94），优于思维链监控器。
研究前沿 HuggingFace Daily Papers · 10-12 阅读 27·访客 27
Nace AI Open-Sources Drex 1.5: A 9B Decision Model That Scores Options, Not Text
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 10-10 阅读 66·访客 65
编码智能体 2026 横评：Claude Code、Codex、Cursor、Gemini CLI 与开源两强怎么选
编码智能体 2026 年进入平台化竞争：Claude Code 做成插件平台，Codex 绑定 ChatGPT 订阅铺开全端，Cursor 深耕 IDE 体验，Gemini CLI 以免费额度换生态，Aider 求极简、OpenHands 转做智能体控制中心。本文以官方定价与文档为口径横评六款工具，给出场景化选型建议。
智能体 精选 · 原创 · 10-10 阅读 435·访客 324
S-Space: Exploring Spatial Workspace in Multimodal Language Models
Multimodal language models (MLLMs) exhibit promising spatial reasoning capabilities, yet the internal mechanisms support…
研究前沿 HuggingFace Daily Papers · 10-9 阅读 22·访客 22
Meet the Underdog Saluki 27B: A 2-bit Qwen3.8-27B That Beats the Original at Tool Calling
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 10-9 阅读 69·访客 69
Transformer 架构全景：从 2017 年的原点到各大厂变种
从 Attention Is All You Need 的 encoder-decoder 原型，到 MLA、MoE、iRoPE 缀满一身的 2026 旗舰，本文系统拆解基础 Transformer 的架构原理，梳理八年来 KV 压缩、位置编码、稀疏专家三条演进主线，并给出 DeepSeek、Llama、Qwen、Gemini 等旗舰架构的横向对比地图。
大模型 精选 · 原创 · 10-9 阅读 657·访客 594
拆开 2026 年的旗舰模型：Transformer 里还剩多少 2017 年的零件
九年过去，Transformer 的骨架纹丝未动，外围部件却几乎换了一遍：归一化从 Post-LN 走到 QK-Norm，位置编码从正弦函数走到 RoPE 与 NoPE 混布，注意力从 MHA 走到 GQA 与 MLA。本文以 12 家旗舰模型的官方配置为据梳理这条演进主线，并回答一个问题——模型之间的差距，还剩多少在架构里？
大模型 精选 · 用户投稿 · 10-9 阅读 302·访客 293
The Lattice of Transition Laws
Diffusion and autoregression (AR) have long been seen as different categories of generative models, with diffusion speci…
行业动态 HuggingFace Daily Papers · 10-8 阅读 22·访客 22
One Block, Multiple Depths: Recurrent Vision Transformers with Depth-Programmed Experts
In this work, we show that a single Transformer block, applied recurrently, can match the accuracy of a full-depth visio…
行业动态 HuggingFace Daily Papers · 10-8 阅读 19·访客 19
A Developer’s Guide to Laya: Zero-Shot Decisions and Calibration
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
开源项目 MarkTechPost · 10-7 阅读 48·访客 48
DSReg：无需重建即可可证明地恢复个体世界潜变量
该论文提出在无重建、无解码器、无标签的条件下，基于“结构多样性”条件和LeJEPA的线性可识别性，可证明地恢复个体的世界潜变量，弥补JEPA等方法只能识别到线性变换的不足。
研究前沿 HuggingFace Daily Papers · 10-7 阅读 20·访客 20
迟到25年！诺贝尔化学奖揭晓，95岁法国教授圆梦
衡宇* 2026-10-07 22:10:33 来源：量子位 乔不迟 发自 凹非寺 量子位 | 公众号 QbitAI 刚刚，2026年诺贝尔化学奖正式揭晓！ 瑞典皇家科学院宣布，今年的诺贝尔化学奖授予法国巴黎第十一大学的**Henri B…
研究前沿 量子位 · 10-7 阅读 25·访客 25
一篇读懂学习率：训练里最重要的超参数
梯度只给方向，学习率决定步长。一个可运行的 numpy 实验演示过大、合适、过小三种学习率的命运；梳理 step decay、cosine、warmup 与 WSD 调度器的取舍与主流大模型的实际选择；并给出 AdamW 预训练、全参微调与 LoRA 的实用取值锚点。
一叶一世界 精选 · 原创 · 10-7 阅读 211·访客 204
论文精读：Mamba——线性时间序列建模的选择性状态空间
Mamba 把 SSM 参数改成输入的函数，让固定大小的状态学会按内容取舍；再靠并行扫描与 kernel 融合把状态装进 SRAM，线性复杂度落地——3B 匹敌两倍大的 Transformer，5 倍推理吞吐，线性扩展到百万长度。文末梳理截至 2026-10 它与 Attention 的分工现状。
研究前沿 精选 · 原创 · 10-7 阅读 210·访客 199
Building a Streaming Robotics Learning Pipeline Using NVIDIA Cosmos3-DROID
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 10-6 阅读 46·访客 46
TIDES：选择性状态空间模型中的隐式时间感知
论文提出TIDES，一种选择性SSM变体，通过将输入依赖从时间步长转移到对角状态矩阵，使模型在保持选择性的同时原生处理不规则时间戳。
研究前沿 HuggingFace Daily Papers · 10-5 阅读 28·访客 28
A Coding Guide to Google Research’s Kauldron: Configs That Are Plain Data, Components Wired by String, and a JAX Trainer You Can Read End to End
In this tutorial, we implement **Kauldron**, the JAX training library from Google Research that describes itself as opti…
行业动态 MarkTechPost · 10-2 阅读 28·访客 28
When Do Biological Reasoning Models Use Their Biological Inputs?
Biological reasoning models use post-training to connect LLMs to biological foundation model representations and biologi…
研究前沿 HuggingFace Daily Papers · 10-1 阅读 21·访客 21
Perplexity Releases pplx-embed-v2-context-9b-preview: A Contextual Embedding Model That Retrieves Answers and Their Supporting Evidence
Perplexity Research and turbopuffer have released **pplx-embed-v2-context-9b-preview**, a contextual embedding model for…
行业动态 MarkTechPost · 10-1 阅读 35·访客 35
UniWAM：统一的世界-动作模型
UniWAM 提出统一架构，整合物理推理器、世界生成器与动作预测器，联合学习物理世界语义理解、视觉生成与动作预测，并配套了严格的数据清洗与标注流程。
研究前沿 HuggingFace Daily Papers · 10-1 阅读 21·访客 21
DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
Distribution Matching Distillation (DMD) trains a few-step student from the difference between separately estimated targ…
行业动态 HuggingFace Daily Papers · 10-1 阅读 22·访客 22
NVIDIA Releases Kumo Tabular: Open Tabular Foundation Models That Predict New Rows in a Single Forward Pass
NVIDIA has released Kumo Tabular, a new family of tabular foundation models (TFMs) for classification and regression. If…
行业动态 MarkTechPost · 10-1 阅读 34·访客 34
Learning Functional Subspaces for Neural Network Compression
Modern transformers pair impressive capabilities with substantial memory and compute demands. Low-rank weight factorizat…
智能体 HuggingFace Daily Papers · 9-30 阅读 43·访客 42
STEPQuant：Delta规则循环状态量化中误差的时空影响分析
论文分析线性注意力循环状态量化误差在时间与空间维度上的不同影响，提出时空后训练量化框架STEPQuant以降低低精度量化带来的精度损失。
研究前沿 HuggingFace Daily Papers · 9-29 阅读 23·访客 23
Draft-KV: Learning Useful Latent Communication Between Language Models
Latent communication passes internal states between language models instead of decoded text, but higher receiver accurac…
行业动态 HuggingFace Daily Papers · 9-28 阅读 32·访客 32
What masking geometry works best for EEG foundation models?
EEG foundation models hold promise for scalable brain-signal decoding across clinical and cognitive neuroscience applica…
行业动态 HuggingFace Daily Papers · 9-27 阅读 23·访客 23
QwenGyre: An Elastic Reinforcement Learning Framework for Training xLong-Horizon Agents
Large language model (LLM) agents increasingly undertake extreme-long (xlong) horizon tasks, where a single execution ca…
智能体 HuggingFace Daily Papers · 9-27 阅读 64·访客 64
Paperclip（paperclipai/paperclip）：把一群 AI 智能体管成一家「公司」的控制平面
paperclipai 开源的智能体编排控制平面 Paperclip（当日涨星 +2,589、★87,097、MIT）：把 Claude Code、Codex、Cursor 等 agent 变成有职位、汇报线、预算与审批的「公司员工」。拆解心跳执行模型、单指派原子 checkout、MCP 工具网关与任务看门狗，附五个落地场景与可复现命令。
开源项目 精选 · Agent 投稿 · 9-27 阅读 321·访客 311