Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs· Flash-dLLM: 针对扩散 LLM 的 I/O 意识缓存和并行解码
Diffusion Large Language Models (dLLMs) have recently emerged as a promising alternative t…
Diffusion Large Language Models (dLLMs) have recently emerged as a promising alternative t…
A multi-agent system can reduce latency on complex tasks by executing work concurrently. S…
Long-term conversational memory in multi-party settings requires more than retrieving rele…
Agents often work on complex problems that require millions of tokens of context, which ne…
We introduce SWE-Serve, a benchmark for evaluating agents on production inference engineer…
Large language model (LLM) agents often handle streams of related tasks, yet standard harn…
Typed decision models are built for settings where model outputs are consumed directly by …
Large language models increasingly tackle hard reasoning problems by spending more test-ti…
A coding agent must emit a valid tool call--a parseable invocation of a tool in the provid…
Successful agent execution need not identify which future product improvement its user wou…
LLM-as-a-judge enables evaluation across diverse tasks, but inference cost and confidence …
LLM agents often use generative models for bounded decisions, raising the question of when…
Pseudorandom number generators (PRNGs) constitute indispensable computational tools across…
AI agents are beginning to automate research and development across the AI stack, from imp…
Adaptive regulation can itself perturb the state it is intended to stabilize. In replicate…
Reliability theory gives a mature language for layered systems, but its formal tools are n…
Learned world models are becoming essential to general-purpose agents: by predicting actio…
Accurately forecasting river runoff is key to managing water resources, controlling floods…
Bot accounts and coordinated influence operations are often discovered via heuristic metho…
A next-token probability says what a model predicts, not how much training support lies be…
Search-based prompt optimizers improve prompts through iterative search: they propose edit…
The rapidly increasing student population has posed serious challenges to the traditional …
We study the multiplayer multi-armed bandit problem with information asymmetry under Berno…
Desktop panel and toolkit for Claude and AI assistants
Desktop companion for AI coding agents and Codex workflows
The incident is the first known breach to affect a government agency, and Australia's prim…
Using GPT-5.6, Ringg powers multilingual agents across voice, chat, WhatsApp, and web for …
9月23日,腾势旗下“科技豪华智能轿车”Z9S正式上市,推出闪充尊荣型、闪充旗舰型、易三方闪充性能型三款车型,面向用户“悦己、悦人、悦非凡”的不同需求。新车搭载第二代刀片电池及闪充…
近日,第八届“儒道心学”心血管病学会议、第十届沪鲁心血管病专家论坛、第九届日照心血管峰会在山东日照召开。由葛均波院士领衔,黄恺、苏国海、李春洁等数十位心血管领域权威专家参与,会上完…
<section data-mpa-powered-by="yiban.io" style="margin-right: 0em;margin-bottom: 0px;margin…
雷峰网获悉,上周,Anthropic启动新一轮合规清查,导致大批AWS和微软大中华区客户的Claude账号被封停。 多名知情人士称,这一轮的封号范围比此前更广。此前外资云中国销售曾…
记者从2026 Intel Connection大会获悉,百度搭子上线面向英特尔机型的本地Skill专区,并推出混合Skill调度能力。用户提出需求后,系统可根据任务类型和电脑算力…
人类演示一次,机器人即可实现跨场景任务复用
<section data-mpa-powered-by="yiban.io" data-pm-slice="0 0 []" style="-webkit-tap-highligh…
9月24日,Accio与数贸会联合推出的专属智能体“数贸会agent”上线,直接开启Agent智能逛展。 本届数贸会参展企业超2000家,光靠腿很难深度逛透。现在只需下载Accio…
专治人机动作对不齐
开源一座具身智能的新“塔台”
懂出海,能记忆,自进化
<section data-mpa-powered-by="yiban.io" data-pm-slice="0 0 []" style="-webkit-tap-highligh…
CEO Mark Zuckerberg kicked off the company’s annual Connect event in Menlo Park on Wednesd…
中科类脑推出新型Token工厂,加速AI发展。
The tiny hardware device creates another mobile home for its AI agent Muse.
Meta says the camera-free glasses will be lighter and have up to 12 hours battery life.
“ <span style="font-family: "Microsoft YaHei Light"; f
But maybe the biggest reveal is that Anthropic has not let Claude run loose in its biology…
The round valued the AI biotech at $2 billion. It is currently testing drugs that treat sk…
Pro and Plus users will be able to use the Work tab on their phones to complete agentic ta…
The report suggests that greater exposure will not resolve the unease around the technolog…
Marking two years of OpenAI Academy and bringing AI skills to even more communities.
YouTube is adding new AI-powered features to YouTube Music that aim to make it easier for …
StrictlyVC joins TechCrunch Disrupt 2026 to discuss the changing VC landscape thanks to AI…
YouTube’s new custom feeds let users describe the videos they want to see in their own wor…
YouTube is adding new features to generate ideas and monitor the performance of thumbnails…
<img data-src="https://static.leiphone.com/uploads/new/images/20260923/6ab3c9f0a4495.jpg" …
OpenAI CEO Sam Altman discusses AI safety, human control, and international cooperation in…
With GPT‑6 Astra, invideo plans edits with greater precision, improves color correction an…
GPT-6 Astra produces more structured, context-aware legal documents, freeing lawyers to fo…
MentalHealthBench is an expert-informed benchmark for evaluating helpful and safe AI respo…
ChatGPT Ads is expanding to Southeast Asia and Taiwan, giving eligible businesses new ways…
Learn how Airbnb is expanding access to GPT-6 Astra and OpenAI frontier models to help eng…

YouTube · Two Minute Papers

YouTube · NVIDIA

YouTube · OpenAI

YouTube · NVIDIA