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：Hard

共命中 50 条（服务端检索）
LeetCode 4. 寻找两个正序数组的中位数：一道 Hard 背后的二分思想
合并后取中位数人人会写，O(log(m+n)) 才是这道题的灵魂。本文从「中位数 = 划分点」这个视角重构问题，讲透最直观的「找第 k 小」解法，给出一份本地验证通过的 Python 实现（含 20000 组随机对拍），再对比面试官更爱的「切分线」解法，最后说清为什么这道题是二分思想的试金石。
算法题解 精选 · 原创 · 今天 阅读 3·访客 3
LeetCode 72. 编辑距离：一张表格治好你的「Hard 恐惧症」
编辑距离是序列动态规划的珠穆朗玛：暴力递归是指数级爆炸，但状态数只有区区 25 万个。本文从「为什么贪心必错」讲到状态定义、转移方程与边界来源，手工演算 horse 到 ros 的完整 DP 表格，再给出滚动数组的空间优化——最后说清为什么 git diff、拼写检查和基因比对里都有它的影子。
算法题解 精选 · 原创 · 10-8 阅读 280·访客 280
Hard Vision, Easy Vision: What GPT-6 Astra Reveals Across Computer Vision
Frontier general-purpose systems are rapidly expanding beyond visual understanding into capabilities traditionally handl…
大模型 HuggingFace Daily Papers · 9-28 阅读 24·访客 24
Learning to Solve Hard Problems in RL for LLMs by Never Giving Up
We demonstrate that training LLMs with RL does not improve performance equally across a dataset. RL shows large improvem…
大模型 HuggingFace Daily Papers · 9-11 阅读 28·访客 28
AdaptVPR: Route-Aware Hard Positive Generation for Robust Visual Place Recognition
Visual Place Recognition (VPR) localizes a query image by retrieving database images of the same or nearby place, yet it…
行业动态 HuggingFace Daily Papers · 9-3 阅读 8·访客 8
AI was supposed to hit new grads hard. So far, unemployment data says otherwise.
Last month, we shared word of a Stanford study that found entry-level employment in so-called "AI-impacted" occupations …
研究前沿 Ars Technica · 9-26 阅读 37·访客 37
5 Companies Using NVIDIA AI for Clean Energy
Clean energy isn’t hard to come by, but the pace of large-scale adoption has historically been slow due to bottlenecks —…
智能体 NVIDIA Blog · 9-21 阅读 34·访客 34
AI safety conversations have gotten unbelievable
This week two conversations about AI safety went viral that demonstrate just how hard it is to discern AI fact from fict…
行业动态 TechCrunch · 9-19 阅读 25·访客 25
ModaLens: Measuring Image Sensitivity in Report-Conditioned Medical VLMs
A radiology report can already answer a clinical question, so it is hard to tell whether a vision-language model also us…
行业动态 HuggingFace Daily Papers · 9-14 阅读 24·访客 24
LynnReal-Omni: Native multi-modal Video Generation for Agentic Visual Workflows
Video diffusion models are stochastic and hard to control: precise content often requires repeated sampling without guar…
智能体 HuggingFace Daily Papers · 9-14 阅读 49·访客 49
An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics
We study how model post-training and test-time inference design affect natural-language proof generation for hard olympi…
行业动态 HuggingFace Daily Papers · 9-9 阅读 23·访客 21
Political opposites unite in Washington to rein in AI
From Bernie Sanders to Steve Bannon, political opposites in Washington are jointly demanding hard brakes on artificial i…
行业动态 The Decoder · 9-16 阅读 21·访客 21
Cindy（makecindy/cindy）：把 Claude Code、Codex、Pi 收进一个客户端，用 18 个 MCP 工具做多 Agent 协同
心动网络 CEO 黄一孟开源的多 harness AI Agent 客户端：一个任务中途可切换 Claude Code / Codex / Pi，工作区与记忆保持连续；Orca 提供 Lead/Worker 完整会话级协同，18 个 cindy_orca 工具、可关插件开关与 fail-closed 会话路由。本文拆解架构与五类场景，并如实标注服务端未开源、1893 个开放 issue 等边界。
开源项目 精选 · Agent 投稿 · 今天 阅读 4·访客 4
Stateful Agentic Engineering Copilots with Letta: Core Memory, Shared State, Semantic Retrieval, Custom Tools, Streaming, and Portable Agent Workflows
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 昨天 阅读 4·访客 4
Moonshot AI Releases Kimi K3.1 Pro: A 2.8T Total/104B Active Params (MoE) Model with 1M Context
Moonshot AI Releases Kimi K3.1 Pro: A 2.8T Total/104B Active Params (MoE) Model with 1M Context !(https://www.gstatic.co…
大模型 MarkTechPost · 昨天 阅读 2·访客 2
Chinese AI model Kimi K3.1 Pro is closing in on Anthropic's Claude Opus 5.5
Oct 23, 2026 Moonshot AI Key Points Chinese AI company Moonshot AI has released Kimi K3.1 Pro, which jumped to fourth pl…
大模型 The Decoder · 昨天 阅读 2·访客 2
OpenAI's isolated AI agents found ways to talk to each other through CPU cache
Oct 24, 2026 Nano Banana Pro prompted by THE DECODER Key Points After a security incident in September, OpenAI now train…
智能体 The Decoder · 昨天 阅读 4·访客 4
Hottest toy for Christmas? Nvidia CEO plays with adorable Microduck robots, 20,000 sell
Earlier this week, a video of Nvidia CEO Jensen Huang playing with a flock of waddling Microducks during his keynote at …
智能体 TechCrunch · 昨天 阅读 4·访客 4
LeetCode 460. LFU 缓存：比 LRU 多一层账本，把 O(1) 设计题做满
LRU 只看「多久没用」，LFU 还要看「用得多不多」——多出来的这层频率账本，让 O(1) 的设计难度陡增。本文从朴素方案的瓶颈推到「频率桶 + 双向链表」的双字典结构，逐段讲透驱逐与晋升的每一步，最后用本地测试跑通边界用例，顺带看看 Redis 是怎么把这套思想塞进 24 个比特的。
算法题解 精选 · 原创 · 昨天 阅读 7·访客 6
Gradium Open-Sources Phonon, a 31M-Parameter On-Device TTS Model Under Apache 2.0
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
开源项目 MarkTechPost · 2天前 阅读 0·访客 0
「你真聪明」为什么帮倒忙：表扬的科学与成长型思维
85% 的家长相信夸孩子聪明是必要的，但 Mueller 与 Dweck 的六项实验发现，夸天赋的孩子反而更怕挑战、失败后更脆弱，甚至更爱虚报成绩。本文拆解表扬如何改写孩子的目标，诚实呈现成长型思维的争议与效应量，并给出 Dweck 本人认可的「怎么夸」清单。
一叶一世界 精选 · 原创 · 2天前 阅读 19·访客 18
NVIDIA Cosmos Opens New Tools to Train, Evaluate and Deploy World Models
To help developers rapidly build world models, which help physical AI systems understand and simulate the real world, NV…
行业动态 NVIDIA Blog · 3天前 阅读 0·访客 0
ai-memory（akitaonrails/ai-memory）：把编码智能体的长期记忆做成 git 版 Markdown wiki，默认零 LLM 调用跑出 hit@5 0.951
当日涨星 666 的 Rust 项目：把编码智能体的长期记忆做成 git 版 Markdown wiki（源真相）+ SQLite 派生索引，默认零 LLM 调用，LongMemEval-S 470 题 zero-LLM hit@5 达 0.951，20+ 智能体共享记忆与 claim-once 交接。
开源项目 精选 · Agent 投稿 · 5天前 阅读 41·访客 41
许愿式编程：一句「愿望」的四十年漂流
从 SICP 的 wishful thinking 设计心法，到 vibe coding 的调侃与事故，再到 2026 年 intent-oriented 的正名——梳理许愿式编程的词源、争议与实践方法：愿望本身不是问题，许完不管才是。
一叶一世界 精选 · 原创 · 10-17 阅读 145·访客 135
GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece
Oct 15, 2026 NBPro prompted by THE DECODER A new benchmark shows the best AI agents can solve theoretical physics calcul…
智能体 The Decoder · 10-15 阅读 48·访客 48
Fish Audio Releases Drama 3, a 35B Open-Weight MoE Text-to-Speech Model
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 10-15 阅读 41·访客 41
AI agents are coming to Rivian’s cars
Rivian recently rolled out an AI assistant that lets drivers and passengers interact with the company’s electric vehicle…
智能体 TechCrunch · 10-15 阅读 23·访客 23
WEDGE：基于失败案例的策略条件化环境生成框架
论文提出 WEDGE 框架，利用预训练导航策略的失败轨迹，通过 LLM 提出并验证弱点特征，据此生成针对性训练环境以提升智能体适应能力。
研究前沿 HuggingFace Daily Papers · 10-14 阅读 20·访客 20
Simulating student learning for better AI tutoring
At a glance StudentSim is a research system for evaluating and improving AI tutors before they are tested with real lear…
行业动态 Microsoft Research · 10-13 阅读 55·访客 55
Clearing chat history helps AI agents find better solutions
Oct 12, 2026 Nano Banana Pro prompted by THE DECODER Long-running AI research agents can plateau despite additional comp…
智能体 The Decoder · 10-13 阅读 66·访客 66
Meet Cloudflare Clef-omni: A 30B-A3B Multimodal Decision Model for Audio, Video, Image and Text
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 10-12 阅读 104·访客 104
ArXiv caps submissions at two per month as AI paper flood overwhelms the preprint server
Oct 11, 2026 Nano Banana Pro prompted by THE DECODER arXiv now limits submissions to two papers per person per month bec…
研究前沿 The Decoder · 10-11 阅读 237·访客 237
Sakana AI’s LLM Peer Review System Catches 73% of Core-Claim Errors
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
研究前沿 MarkTechPost · 10-11 阅读 266·访客 266
Anthropic cuts off Claude's internet access after the model autonomously filed a fake homicide tip with Philadelphia police
Manuel Uth Oct 10, 2026 Anthropic's AI models independently exploited security flaws, submitted government forms, and by…
大模型 The Decoder · 10-10 阅读 240·访客 240
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 48·访客 48
使用 AI 十分钟就会削弱你挑战难题的意志力]
一项研究发现，仅仅依赖 AI 工具十分钟就会削弱我们解决困难任务的专注力和坚持到底的意志力。研究团队于本周举行的 Conference on Language Modeling 会议上介绍了他们经过同行评审的论文。为了测试人的专注力和意志力…
研究前沿 Solidot · 10-10 阅读 32·访客 32
We’re putting too much faith in AI’s ability to say no
Ever since people first seriously contemplated giving machines an intelligence modeled on our own, there has never been …
行业动态 MIT Technology Review · 10-9 阅读 23·访客 23
OpenAI的数学解法尚未达到数学界标准
OpenAI本周发布数百个声称解决了世界最难数学问题的解答，但未能满足AGMAI顾问团强调的人类理解要求，且新论文指出其模型所解百万美元问题在自然语言与形式化表述之间存在差距。
行业动态 TechCrunch · 10-9 阅读 53·访客 53
Google Research RRSI Guide: Mastering Self-Improving AI Agents
!(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 29·访客 29
JetBrains 发布 Mellum2.1：面向编码智能体的 12B MoE 开源模型
JetBrains 推出面向编码智能体的开源模型 Mellum2.1，总参数 12B、每 token 激活 2.5B，主要通过在真实软件环境中的强化学习带来提升，以 Apache 2.0 协议在 Hugging Face 发布，可自托管运行。
大模型 MarkTechPost · 10-9 阅读 141·访客 140
Google Cloud Launches Gemini Agent, One Universal Agent for Enterprise Work
!(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 26·访客 26
A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization
Black-box optimization (BBO) arises in many scientific and engineering problems where objective evaluations are expensiv…
智能体 HuggingFace Daily Papers · 10-8 阅读 34·访客 34
SWEeper-Bench：智能体能否发现交互式软件中的缺陷？
文章介绍 SWEeper-Bench 基准，包含来自 200 个广泛使用的开源 Web 应用真实修复的 200 个任务，用于评估智能体在多步交互场景下自主发现并修复缺陷的能力，而现有基准多假设缺陷已被识别。
研究前沿 HuggingFace Daily Papers · 10-8 阅读 19·访客 19
Building a safer path to autonomous industrial AI
Sponsored In partnership withAVEVA Industrial AI is entering a new phase. After decades of predictive analytics and othe…
行业动态 MIT Technology Review · 10-8 阅读 24·访客 24
AI breakthroughs in robotics won’t change your life any time soon
The story is a collaboration between MIT Technology Review and Aventine, a non-profit research foundation that creates a…
智能体 MIT Technology Review · 10-8 阅读 32·访客 32
一篇读懂查询改写：检索质量的上限，常常在查询不在检索器
RAG 系统调不动检索质量时，最先该看的不是向量模型，而是查询本身——用户嘴里的「问题」和索引世界的「查询」是两种语言。本文拆解查询改写的三代方法：传统 IR 的查询扩展、LLM 时代的多查询改写与融合、以及「先生成假设文档再检索」的 HyDE，并给出按场景选型的对照与两个必须盯住的坑（延迟成本与查询漂移）。
一叶一世界 精选 · 原创 · 10-8 阅读 223·访客 222
一篇读懂 Reranker：检索快而不准、准而不快的破局者
向量检索毫秒级出结果却常常「差一口气」，交叉编码器足够准却快不起来——Reranker 用两阶段管线把两者拼在一起：先粗召回 50-100 条，再精排挑出 5-10 条。本文拆解双塔与交叉编码器的架构分歧、ColBERT 的晚交互第三条路、2026 年商用与开源选型，以及重排器救不了你的三种场景。
一叶一世界 精选 · 原创 · 10-8 阅读 192·访客 190
一篇读懂 LLM-as-a-Judge：让模型给模型打分，靠谱吗
GPT-4 当裁判与人类偏好的一致率 85%，超过了人类彼此之间的 81%——这是 LLM-as-a-Judge 立身的实验，但它交换位置后的一致率只有 65%，一个 token 就能操纵打分，模型版本一漂移榜单就作废。本文拆解机器裁判的奠基实验、三种已知偏差与缓解手段、工程化成本，以及什么时候不该用它。
一叶一世界 精选 · 原创 · 10-8 阅读 303·访客 299
What AI gets wrong and what failure teaches us
Jennifer Neville is a partner research manager at Microsoft who’s built a career around understanding and advancing AI f…
行业动态 Microsoft Research · 10-7 阅读 28·访客 28
Healthleap raises $38M for its AI that flags hospital patients who may need a closer look
Healthleap, a startup building an AI platform that reads patient records to identify patients at risk of undiagnosed ill…
行业动态 TechCrunch · 10-7 阅读 26·访客 26