本周 AI 热榜

Nvidia CEO Jensen Huang resigns amid spy, affair scandals

Google Releases Gemini 4 Hydrogen: 2.5x Faster Than Gemini 3.8 Flash

AI unlocks Gödel's notebooks, said to contain hundreds of proved theorems

ChatGPT’s new Wallet lets AI agents shop and check out for you

Jensen Huang’s departure blasts big hole in AI startup funding, VCs and founders say

Anthropic's IPO is six times oversubscribed and on track to become the largest public offering in history

Nvidia’s stock drops as board reportedly preps to oust CEO Jensen Huang

OpenAI rebuilt AI shopping from scratch with ChatGPT Wallet

Best AI Video Generation Models in 2026: Elo, Price per Second, Open Weights

Google Releases TPU Sync: Open-Source KV Cache Transfer for TPUs

Former OpenAI researcher says he found Nvidia Blackwell fingerprints hidden inside six Chinese AI models

Jonathan Blow – If AI solved coding, why is my software still the same? [video]

Introducing ChatGPT Wallet

Jensen Huang is out at Nvidia — who is Colette Kress, his interim replacement?

Amazon has a plan to index the world’s online shopping data

llama.cpp v1.0.0 : A stronger foundation for the local AI ecosystem

Production MLOps at GS Caltex with Amazon SageMaker Canvas, part 1

Route coding agent requests to the right model on Amazon Bedrock with NVIDIA NeMo Switchyard

Architecture patterns for fast session startup on Amazon Bedrock AgentCore runtime

OpenFold3 on Amazon SageMaker JumpStart: Predict biomolecular structures

AI is destroying the trust between writers and readers

Jensen Huang's ouster likely traces back to Nvidia's risky business practices more than his personal life

No AI Slop: A Manifesto

AI agent startup LangChain hits $3B valuation

Cost-optimizing AssemblyAI speech-to-text inference on AWS with Amazon EC2 Spot

Introducing Gemini 4 Hydrogen

Cost attribution and controls for AI developer and productivity tools on Amazon Bedrock

Export-banned Nvidia chips are behind China’s most powerful AI models — uh-oh, Jensen!

Google Deepmind's AlphaProof Code verified Signal's cryptographic core in four weeks instead of years

Google’s new Gemini model is better at understanding complicated questions

Google Deepmind shifts genomics AI from bulk tissue predictions to single cell type resolution with AlphaGenome 2

Banks bar Chinese and Hong Kong investors from Anthropic IPO over US export controls

Build a model improvement flywheel with Amazon SageMaker HyperPod Data Capture

Build a Robot Pick-and-Place Simulation with SimReady Assets

Edge0-145B: Frontier AI Capabilities on Your Own Computer

Read Nvidia’s internal emails about the ‘extraordinarily difficult period’ amid CEO shakeup

OpenAI board member worries an AI "takeover" of society could be months away

XYBench: Helping with the problem behind the question

Chinese frontier AI models carry the fingerprint of export-controlled GPUs

Anthropic's Claude for Manufacturing plugs AI directly into factory software to automate industrial workflows

Nearly 70 percent of US agencies use generative AI, and unions are finally catching up with guardrails

OpenAI enters the cybersecurity market with two Daybreak products

Back to Pixels: Lifting Latent Video Diffusion into Pixel Space

Classical Energies Learn to Reason 1,000x Faster than Looped Models

Goal Encoding in LLMs

What’s really at stake in the Jensen Huang saga

ElevenLabs’ new app combines meeting note-taking and dictation for teams

DeepSeek releases DSpark - 50%-600% faster spec decoding vs MTP

DeepSeek just released DSpark for V4 Flash & Pro, a new speculative decoding

Claude’s Duet agent is a little neurotic and surprisingly helpful

OpenAI’s data suggests that frontier models are hiding their thoughts because developers optimize for capability and compute

Your Vision Foundation Model Is Secretly an EEG Foundation Model

How AI Is Helping Care Teams Get Patients the Right Care, Sooner

United Airlines bans AI agents from booking flights on behalf of travelers

Gemini 4 Carbon is now available

Why You Should Expect an AI Crash

aws-emulator

Anthropic targets $1.4 trillion IPO

Anthropic sets IPO price range of $70 to $78 per share to reach an implied valuation of $1.404 trillion

China’s Internet Giants Are Listing AI Units Because the Parent Stock Will Not Price Them

How Natera designed an agentic AI system for lab ops with Amazon Bedrock AgentCore

Elevated errors on claude.ai, Claude Code, and platform.claude.com

Reins: Frontier Multimodal Models Driving Real Robot Arms on Long-Horizon Tasks, Zero-Shot

Singularity-Bench: Can Agents Build Agents?

Claude Meetings lets Anthropic's AI join your video calls

ZONOS3-Mini Technical Report

Google's Gemini 4 Hydrogen outscores the bigger Argon just four weeks after its launch

Google’s AlphaGenome 2 zooms in on the cellular roots of disease

Google rolls out Gemini 4 Hydrogen, finally replacing 3.8 Flash

Unbundling Multimodal Vision Encoding

K2V: Generative Keyframing For Physically Plausible Video Generation

Learning When Diffusion Language Models Should Regenerate

Linking Rapid Learning to Compositional Structure in Neural Networks: Evidence from Additive Analogies in Meta-Learned Parameter Spaces

Consistent Video-to-Video Translation via Explicit Correspondences

LiveGeo: Benchmarking Geographic Agents in Continuously Evolving Worlds

Optimizing cost per task and proving agent ROI on Amazon Bedrock

BlinkMimic: Training a Humanoid Motion Tracker in Minutes on a Single GPU

Ballad: Bandit-Based LLM Routing for Automated Heuristic Discovery

Objects as Addresses: Instance Binding in Vision Transformer Attention

Google Search gets a major upgrade with Gemini 4 Hydrogen

Maximizing Capacity Utilization of LoRA Adapters for Mitigating the Learning-Forgetting Trade-off

Neural Render Proxies for Interactive and Differentiable Lighting

SpotDiff: A Plug-and-Play Framework for Zero-Shot Change Detection in Remote Sensing Imagery

Actor, Judge, or Teacher? Understanding Decision Models in Long-Horizon Agents

XBot-Omni: An Omni-Modal Model for Embodied Understanding and Interaction

AWS Releases AWS Emulator: A Local Copy of 200+ AWS Services Built by AI Agents

STeP-C: Real-Time Neural Compression for Distributed Fiber-Optic Sensing at Scale

Preserve the Vulnerability, Break the Alert: Testing the Robustness of CodeQL Taint Tracking with Natural Rewrites

CapLite: Prompt-Injection Defense for Small Language Model Agents without Model-Written Security Artifacts

SALA: Semantically Aligned Latent Actions Enable Dual-Axis Scaling of Embodied Foundation Models

Spatiotemporal learning improves quantification of pore-space evolution and fluid displacement

LazyWeights: Decode-on-Demand Edge LLMs

From prototype to production: Engineering agents with Agent DLC

Nvidia CEO Jensen Huang has resigned. What happens next?

Google pushes privacy angle as Gemini-powered smartglasses debut starting at $395

Chinese Workers Are Paying Their Own AI Bills to Stay Productive

OpenAI reportedly raises an additional $11 billion at a $1.25 trillion valuation

Cheap AI Tracks Are Crowding Human Arrangers Out of China’s Playlist Economy

Anthropic's Fable 5.2 completes 41 percent of real freelance jobs, up from 2.5 percent a year ago

Anthropic kicks off IPO roadshow, offering shares at $70 to $78 and targeting a trillion-dollar valuation

Introducing AlphaProof Code: a breakthrough in verifying complex software at depth

FAA expands AI advice to manage air traffic congestion across entire US airspace

Darmok: Championship Cryptic Crossword Solving via Constraint Tightening and Resilient Orchestration

IBM Announces IBM Quantum Nighthawk r3: A New Processor to Support 10k Quantum Operations on 520 Programmable Qubits

OpenAI’s quarterly revenue reportedly crosses $10B for the first time

Claude Fable 5.2 and Mythos 5.2 cement Anthropic's AI lead just before its IPO

Meta Releases Muse Ember, a 1M-Context Model for High-Volume Agent Work

Loyalty Capture: Reporting Relationships and Structural Sycophancy in Frontier AI Models

Liquid AI Releases LiquidCodec, a 74M-Parameter Streaming Audio Codec

Budget3D Image-to-3D Generation at Any Token Budget HF: https://huggingface.co/BudgetAny

Complex systems fail suddenly and catastrophically for complex reasons.

extra salty

Check out our recent results solving open challenges across math/science and engineering with /teamwork and Gemini 4 models : http://goo.gle/3XpyJNG D...

Today, we are sharing new breakthroughs from our Antigravity’s /teamwork multi-agent systems. Through ongoing collaboration with academic partners, o...

Holy shit

🛳️🛳️🛳️

20 hours later, could Will confirm that V4.1F was trained on Nvidia?

V4.1 Flash is in good company, wow

we are launching our latest decision models, pplx-decider-v1.3 and pplx-decider-flash-v1.3. for this release we prioritized robustness, so the models ...

G-STAR Group-Scaled Token Capacity for Deeply Compressed Autoregressive Generation paper: https://huggingface.co/papers/2611.13004

Decision Models are the future of agentic security. They can process billions of tokens of telemetry, identify security issues, and help you patch fix...

NVIDIA doesn’t own a fab. Its edge is a supply chain that took one product from $1 billion to $250 billion. Even so, it will probably lose share. “W...

T3 Code now lets you fully customize your UI in app. Ask your agent to make changes and it will update on the fly

🚨 Our Generative AI Lab is releasing Primer: an agent-driven system where a teacher talks through lessons & gets back a playable simulation for the...

🎉 Day-0 support for @thinkymachines’ Inkling-2-Small in vLLM! A 500B MoE with text + image inputs and text output. Bring the image into the conver...

The legendary tonbi has made a plugin :O

Baidu's ERNIE-5.5-Flash-Preview is now usable for common, moderately difficult agent tasks. But in tests by Zhihu contributor toyama nao, reliability ...

Introducing GR00T 2: our most advanced, open System One model for robotics. It runs entirely on your robot’s onboard compute, fully untethered, and y...

one of the capabilities we are pushing very hard on for the decider model is security. we've already found many internal applications for it, and we n...

Formalising CPython in Lean with Opus 5.5, finding a 5 line segfault crash

Very impressed by this. In the harness world, you often don't get the performance you need when you try to maximize efficiency and speed. Looks like R...