Kimi Paper Reveals Inference Architecture, Handling 80% of Traffic
Kimi paper reveals inference architecture handling 80% of traffic: AI forward-looking topic (batch3, outside main timeline Jan 2025–May 2026).
Topics
Browse AI news across models, agents, media, industry, and compute policy.
Filter by category
Llama Can Now Generate Images: HKU and ByteDance Release Open-Source Autoregressive Text-to-Image Model
HKU and ByteDance have released an open-source autoregressive text-to-image model, enabling Llama to generate images. Online demos are now available.
Report: AI Unicorn StepFun's Valuation Doubles in Six Months, Reshaping the 'Big Six' Landscape
StepFun's valuation has doubled in six months, shifting the large model startup landscape into a new 'Big Six' configuration.
Talent from Yao Class Teams Up for Sequel to 'Oh No! I'm Surrounded by Large Models'!
Yao Class prodigies collaborate on the sequel to the hit game, introducing AI tools for workplace efficiency.
Kaiming He's New AI Generation Project: First Team Leadership at MIT, Featuring Double Olympiad Gold Medalist Mingyang Deng
Kaiming He leads his first team post-MIT hiring on a new AI generation project, with double Olympiad gold medalist Mingyang Deng participating. This work falls outside the main 2025-2026 publication timeline.
The Robot World Jensen Huang Envisions Still Needs AI Data for 'Training' | CVPR 2024
Jensen Huang's vision of a robot world still requires AI data to be trained. This is an AI forward-looking topic (batch 3, outside the main timeline of Jan 2025–May 2026).
Surpassing Transformers: Tsinghua and Ant Group's Pure MLP Architecture Boosts Short- and Long-Term Time Series Forecasting
A new pure MLP architecture from Tsinghua University and Ant Group outperforms Transformers in short- and long-term time series forecasting.
Domestic Generative AI Model for Antibody Design Matches AlphaFold3, Predicts Antigen-Antibody Complex Structures and Enables De Novo Antibody Design
A new domestic generative AI model matches AlphaFold3 in performance, simultaneously predicting antigen-antibody complex structures and enabling de novo antibody design.
Surpassing Devin: Yao Class Team Sets New Benchmark in Large Model Programming
Surpassing Devin! Led by the Yao Class, they set a new record in large model programming. (AI Forward Agenda, batch 3; published outside the main timeline of Jan 2025–May 2026).
BaiChuan's New Model Tops Chinese Benchmark; First AI Assistant 'Bai Xiaoying' Launched as 'Most Search-Savvy'
BaiChuan unveils a new model that tops Chinese benchmarks alongside its first AI assistant, Bai Xiaoying, dubbed the most search-savvy. This falls outside the main 2025-2026 timeline.