搜索:Deep

共命中 50 条(服务端检索)
Deep Research 类智能体:研究型 Agent 的架构拆解与产出评估
OpenAI Deep Research、Gemini Deep Research 等研究型智能体把「深度调研报告」变成了分钟级交付。本篇拆解这类 Agent 的架构共性(多轮检索规划、广度与深度的平衡、交叉验证)、与单次搜索问答的本质差异、产出的真实质量分布,以及怎么用与怎么评估这类「研究助理」。
智能体 精选 · 原创 · 11-4 阅读 61·访客 61
从搜索框到 Deep Research:Agentic 检索怎么把一次查询变成一场调查
Google 比 OpenAI 早七周发布 Deep Research,但把「研究计划让用户改批」产品化的也是它——agentic 检索的通用循环是:规划、迭代检索、阅读、反思补漏、交叉验证、带引用报告。本文拆解这个循环的两种实现路线(端到端 RL vs 显式编排),用 BrowseComp 上「裸模型不足 10% vs deep research 51.5%」的差距说明多步浏览行为本身值多少分,也把「慢不等于对」的引用可靠性研究摆上台面。
智能体 精选 · 原创 · 10-8 阅读 674·访客 666
Beyond Watching and Searching: Benchmarking Agents for Video-centered Deep Research Reports
Deep research agents synthesize open-web evidence into long reports, but current benchmarks do not evaluate research cen…
智能体 HuggingFace Daily Papers · 10-22 阅读 76·访客 76
Deep Search in Real-World Web Environments
Despite substantial progress in Deep Search agents, their ability to conduct end-to-end research in real-world open Web …
智能体 HuggingFace Daily Papers · 10-19 阅读 58·访客 58
OneSearch-VL: Unified Multimodal Deep Research Agent for Image and Video
Single-image, multi-image, and video deep research require different visual operations but share a workflow of visual gr…
智能体 HuggingFace Daily Papers · 10-8 阅读 88·访客 87
Incremental Open-Ended Deep Research with Structured Harness
Existing Open-Ended Deep Research (OEDR) systems primarily generate reports from scratch, making them inefficient for sc…
行业动态 HuggingFace Daily Papers · 10-8 阅读 28·访客 28
IterSynth: Rethinking Deep Search Agents via Role-Decoupled Iterative Synthesis
Deep search requires LLM agents to decompose complex queries, search for evidence, and synthesize grounded answers, yet …
智能体 HuggingFace Daily Papers · 9-24 阅读 89·访客 88
Deep Principle Releases Hydrogen-1.0: A 753B Open-Weight Model for Scientific Discovery
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 11-3 阅读 33·访客 33
Stochastic Deep Supervision for Gallery-Consistent Early-Exit Face Recognition
Vision Transformers (ViTs) perform strongly in face recognition (FR), yet their fixed-depth inference limits deployment …
行业动态 HuggingFace Daily Papers · 11-2 阅读 15·访客 15
DR Pixel: Page-Native Deep Research with Vision-Language Models
Complex documents such as scientific papers contain rich information in tables, figures, and equations; however, existin…
智能体 HuggingFace Daily Papers · 11-2 阅读 20·访客 20
Deep Fusion Memory: A Chunk-Level Parametric Memory for Large Language Models
Large language models need effective ways to acquire and use external knowledge beyond what they retain from pretraining…
行业动态 HuggingFace Daily Papers · 10-18 阅读 22·访客 22
OmniCapBench: A Deep-Structured Evaluation Framework for Fine-Grained Audio-Visual Captioning
Multimodal large language models (MLLMs) are rapidly evolving toward continuous audio--visual reasoning, creating an urg…
研究前沿 HuggingFace Daily Papers · 10-8 阅读 29·访客 29
Exa Launches Agent Ultra: A Subagent Swarm Deep Research API Built for Exhaustive List Building
Exa has released **Agent Ultra**, the highest effort level of its Exa Agent API. It is built for research that must run …
智能体 MarkTechPost · 9-26 阅读 113·访客 112
Sakana AI hires Jürgen Schmidhuber, inventor of deep learning, world models, and your next ChatGPT update
Sep 24, 2026 Silicon Valley's next big AI idea is probably already sitting in a 1991 paper by Sakana AI's new chief advi…
研究前沿 The Decoder · 9-25 阅读 50·访客 50
Deep Persona: A Psychologically Grounded Architecture and Evaluation Framework for Role-Playing Agents and Simulations
Existing approaches to persona simulation with Large Language Models (LLMs) mostly rely on shallow character description…
智能体 HuggingFace Daily Papers · 9-6 阅读 88·访客 87
DeepStress: Stress-Testing Deep Search Agents
While search agents demonstrate impressive capabilities in multi-step question answering, their robustness to poor-quali…
智能体 HuggingFace Daily Papers · 11-12 阅读 8·访客 8
英国水果蔬菜营养价值下降最高五成]
Deep Science Ventures 研究人员的一项研究发现,英国水果和蔬菜的营养价值自 1960 年代以来减少最高五成。英国民众现在需要食用两倍数量的蔬菜,才能摄入与 60 年前同等水平的矿物质如钾、镁或铜。许多农作物中几种关键营养…
研究前沿 Solidot · 10-11 阅读 204·访客 204
AI 做数学研究走到了哪一步:奖牌、轰炸与「数学 2.0」之争
从 IMO 奖牌级表现到一次发布 372 项结果引爆抵制,AI 证明在 2026 年接连突破也接连越界。本文梳理 AlphaProof、Gemini Deep Think 与 OpenAI 最新成果、Lean 形式化生态的进展,厘清机器辅助证明当前的能力边界,以及数学社区围绕「数学 2.0」的核心争论。
研究前沿 精选 · 原创 · 10-10 阅读 760·访客 697
Not All Objectives Are Born Equal: Priority-Constrained Descent for Hierarchical Multi-Objective Optimization
Deep learning problems rarely involve objectives that are equal in importance. A primary objective defines the goal, whi…
行业动态 HuggingFace Daily Papers · 9-21 阅读 30·访客 30
StepAudio 3 Realtime Technical Report
Realtime spoken interaction demands deep reasoning, prompt responses, and fluid turn-taking. We present StepAudio 3 Real…
研究前沿 HuggingFace Daily Papers · 9-12 阅读 45·访客 45
MaxKernel: Agentic Kernel Generation for TPUs
Designing and authoring high-performance custom kernels for accelerators is a complex task that requires deep hardware-l…
智能体 HuggingFace Daily Papers · 9-3 阅读 90·访客 89
ShallowStream: Index Shallow then Answer Deep for Streaming Video Understanding
Streaming video understanding is a critical capability for real-world applications, including embodied intelligence, aut…
行业动态 HuggingFace Daily Papers · 9-2 阅读 32·访客 32
给定骨架再写序列:逆折叠设计能做到哪一步
逆折叠把目标骨架作为条件来生成氨基酸序列,但可计算的序列恢复和候选设计不等于实验中的折叠、装配或功能。本文比较 ESM-IF1 与 ProteinMPNN 的输入输出及验证边界。
研究前沿 精选 · 原创 · 昨天 阅读 11·访客 11
分子筛选中的不确定性:把模型分歧变成可校准的决策
分子预测里的“高不确定”不是一种单一现象:集成模型分歧、真实标签噪声、预测区间校准、分布移位与筛选决策各回答不同问题。结合分子性质不确定性比较、evidential learning、原子级归因和 DataSAIL,解释如何校准、如何识别评测泄漏,以及何时应追加实验、改变优先级或拒绝自动筛选。
研究前沿 精选 · 原创 · 昨天 阅读 13·访客 13
从序列中学到什么:蛋白质语言模型的表示学习,不是蛋白设计
蛋白质语言模型先从序列中学习上下文表示,再由下游任务读取其中的信息。本文梳理掩码预训练、结构与突变效应评测,并说明表示学习与生成式设计之间的边界。
研究前沿 精选 · 原创 · 昨天 阅读 12·访客 12
从单步拆解到整条路线:AI 逆合成的搜索逻辑
单步逆合成模型只回答目标分子下一步可能怎样拆,多步规划器则要把候选拆解组成通向可购起始原料的路线。以 Segler 的 MCTS 与 Retro* 为例,说明搜索如何引导、评测能证明什么,以及为什么规划结果不等于实验验证。
研究前沿 精选 · 原创 · 昨天 阅读 3·访客 3
原子势的主动学习:把昂贵的 DFT 标签用在最需要的位置
训练原子势时,盲目增加随机结构不一定改善目标轨迹。主动学习从模型分歧、偏差或任务相关价值中挑出待标注构型,再用 DFT 回填数据并迭代训练。本文拆解探索—筛选—标注闭环,解释不确定性采样的盲点,并结合 DP-GEN 与材料科学主动学习综述,给出暂停、校准和审计建议。
研究前沿 精选 · 原创 · 昨天 阅读 2·访客 2
AI 科研的可复现性与不确定度报告:让结果能够被检查
科学 AI 的一个分数或一条模拟轨迹,只有在数据、代码、模型版本、计算设置和误差来源可追溯时才有解释力。本文区分 National Academies 定义下的计算可复现与独立复现,介绍 FAIR 数据原则和深度网络不确定度分类,并以 MatterSim 官方说明为例,讨论开源权重、模型适用域与外部验证之间的差异。
研究前沿 精选 · 原创 · 昨天 阅读 3·访客 3
Clearing Up the Confusion on AI Data Retention: Two Standard Designs Explained
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 4天前 阅读 28·访客 27
OpenAI Releases GPT-6.2 Astra and Sol With a 20% Cheaper Astra API
!(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 27·访客 27
The US wants to block China from renting AI power abroad but China already built a parallel supply chain from US AI output
Nov 17, 2026 Nano Banana Pro prompted by THE DECODER Chinese companies have been skirting US export bans by legally rent…
行业动态 The Decoder · 4天前 阅读 21·访客 21
The Download: Jensen Huang’s open-source dilemma and cognitive immortality
This is today's edition of* *The Download*,*our weekday newsletter that provides a daily dose of what's going on in the …
开源项目 MIT Technology Review · 5天前 阅读 20·访客 20
How the World’s Most Advanced AI Infrastructure Supply Chain Invents the Future
Deb Shoquist leads operations at NVIDIA, overseeing the supply chain and manufacturing behind its AI platforms. That des…
行业动态 NVIDIA Blog · 5天前 阅读 20·访客 20
Liquid AI Releases d1.2 and d1.2-ultra: Omnimodal Decision Models That Top the AA Decision Index
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 5天前 阅读 8·访客 8
The open-source movement wants Jensen Huang back
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox fir…
开源项目 MIT Technology Review · 5天前 阅读 30·访客 30
Alibaba's Qwen4-Max takes on Opus 5.5, GPT-6, and Kimi K3.1 Pro at much lower prices
Nov 16, 2026 Key Points Alibaba launched Qwen4-Max, whose revamped architecture processes long texts more efficiently an…
大模型 The Decoder · 5天前 阅读 32·访客 32
Meet the new James Bond: Robert Pattinson
The Internet has been abuzz for months about which actor would replace Daniel Craig as British agent 007 James Bond in t…
智能体 Ars Technica · 5天前 阅读 14·访客 14
Building trustworthy AI by fencing it in, not predicting it
Sponsored Provided byUL Solutions In October this year, OpenAI disclosed that two of its AI agents had switched off the …
智能体 MIT Technology Review · 5天前 阅读 15·访客 15
Sarvam Launches Sarvam Token Factory: Bringing India-Hosted Inference to 8 Open Models
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 5天前 阅读 8·访客 8
NVIDIA Open-Sources NVKE and NVKR for Managed Kubernetes on GPU Infrastructure
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
开源项目 MarkTechPost · 5天前 阅读 10·访客 10
CUDA at 20: How the Platform Transformed Scientific Computing
This week at SC26, NVIDIA is marking two decades of CUDA by honoring the researchers, developers and engineers who helpe…
行业动态 NVIDIA Blog · 5天前 阅读 7·访客 7
AI Day Seoul Brings Together Korea’s Industry Leaders, Startups and Developer Communities
Over two days at COEX, more than 2,000 attendees converged for AI Day Seoul, bringing together Korea’s industry leaders,…
智能体 NVIDIA Blog · 6天前 阅读 6·访客 6
Alibaba Releases Qwen4-Max: 2.4T MoE, 71B Active, 1M Context
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
大模型 MarkTechPost · 6天前 阅读 31·访客 30
K2-Horizon: A Frontier Fleet of Open Source Language Models
We present K2-Horizon, a fleet of six fully open language models, from a 0.9B dense model to a 375B-A23B MoE, built for …
智能体 HuggingFace Daily Papers · 6天前 阅读 7·访客 7
Meituan releases a 1.6-trillion-parameter agent model that never touched an Nvidia GPU
Nov 15, 2026 ChatGPT-5.6 prompted by THE DECODER Meituan releases the weights for LongCat-2.5, a 1.6-trillion-parameter …
智能体 The Decoder · 11-15 阅读 33·访客 33
Cmpnd Releases dspy-wsp: Weight-Space Programming Comes to DSPy
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 11-15 阅读 26·访客 26
Perplexity Releases Decider v1.3: Open-Weight Decision Models Add Locate
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 11-14 阅读 29·访客 29
Stanford Research Open-Sources Memorilla, a Latent Memory Module for Frozen LLMs
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
开源项目 MarkTechPost · 11-14 阅读 33·访客 33
Nvidia privately confirmed it could track chips to fight smuggling, even as industry publicly questioned feasibility
Representatives from Nvidia told Illinois representative Bill Foster that it would be technically possible to implement …
行业动态 MIT Technology Review · 11-13 阅读 11·访客 11
Liquid AI Releases LiquidCodec, a 74M-Parameter Streaming Audio Codec
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 11-13 阅读 14·访客 14