At the 2025 CCF Global High Performance Computing Conference, Researcher Lin Pengfei from the Institute of Atmospheric Physics, Chinese Academy of Sciences, presented a keynote speech titled "A Deep Learning Model (CoTCN) for Improving the Accuracy of Short-Term Global Sea Surface Temperature Forecasts." This article contains a transcript of the highlights of Lin's presentation.

Anthropic CTO Tom Brown reflects on his transition from entrepreneurship to AI research.

HyperAI has compiled a series of valuable and widely used tutorials and datasets from August 25th to 29th, covering a variety of fields such as image generation, visual mathematical reasoning, and more.

To address the issue of coastal ecosystem health diagnosis, a team from the Hong Kong University of Science and Technology proposed the Spatiotemporal Interpolation and Prediction (STIMP) model, which achieves accurate prediction of the spatiotemporal distribution of chlorophyll a by integrating specially designed modules.

8.25-8.29 AI Paper Selection

"Deploying NVIDIA-Nemotron-Nano-9B-v2 with vLLM + Open WebUI" is now available in the "Tutorials" section of the HyperAI Hyperneuron website (hyper.ai). Come and experience communicating with this "small but precise" large language model!

Recently, Xiaohongshu hi lab released a multilingual document layout parsing model called dots.ocr, which once again refreshed the height of OCR technology with its 1.7B lightweight design and SOTA-level performance.

At the 2025CCF National High Performance Computing Academic Conference, Researcher Zhang Zhengde, head of AI4S at the Computing Center of the Institute of High Energy Physics, gave an in-depth presentation on how intelligent agent technology can be applied to the processing of AI-Ready data.

A research team from the Department of Chemical Engineering and Applied Chemistry at the University of Toronto, Canada, proposed a multimodal machine learning approach to identify MOFs that have potential in areas different from their originally reported applications.

HyperAI has compiled a series of valuable and widely used tutorials and datasets from August 18th to 22nd, covering areas such as image editing and multimodal reasoning.

NVIDIA Physical AI has made another breakthrough, with the robot brain going online, enabling real-time intelligent interaction with the physical world.

The MIT research team combined chemical informatics tools with the new organic solubility database BigSolDB, enabling the model to simultaneously input solute molecules, solvent molecules and temperature parameters, and directly perform regression training on logS.

Li Mingchen, a postdoctoral fellow in Hong Liang's research group at the Institute of Natural Sciences, Shanghai Jiao Tong University, shared the latest research progress and technological breakthroughs in the basic model of proteins and genomes with everyone under the theme of "Proteins and Genome Basic Models".

August 11-15 AI Paper Selection
Contains high-quality data set download address

The "VenusFactory Protein Engineering Design Platform" has been launched on the tutorial section of HyperAI's official website (hyper.ai), covering 7 functional modules and can be experienced online with one-click deployment.

Professor Zhuang Yingping from East China University of Science and Technology shared her views on "AI Assisting Efficient Biomanufacturing Processes" and introduced the technical system and team achievements from three aspects, including the relationship between biomanufacturing and synthetic biology.

A team from Cornell University has proposed an integrated circuit called a microwave neural network (MNN) that can simultaneously process ultra-high-speed data and wireless communication signals, providing a new solution for high-bandwidth applications.

8.11-8.15
Weekly Highlights!

Contains 8 Qwen online tutorials

Medical GraphRAG effectively improves the credibility and overall quality of answers by generating evidence-based answers and official medical terminology explanations.

August 11-15 AI Paper Selection

Ainnova Tech, a health technology company, has created the Vision AI platform, which can detect diabetic retinopathy (with an accuracy rate exceeding 90%), cardiovascular risk, and other multi-system diseases in seconds.

The UCLA research team proposed a virtual histological staining method based on a diffusion model, which enables the prediction of high-resolution cell tissue pathological structure based on low-resolution IMS data.

The Alibaba Tongyi Qianwen team has once again open-sourced its first basic image generation model, Qwen-Image, with a parameter count of up to 20B, refreshing the SOTA state-of-the-art image generation model.

Paper With Code has officially ceased operations, and deep users around the world have spoken out.

"MediCLIP: Anomaly Detection in Small Sample Medical Images Using CLIP" is now available in the "Tutorials" section of the HyperAI website. Come and experience this efficient method for intelligent medical imaging diagnosis!

HyperAI has compiled a series of valuable and widely used tutorials and datasets from August 4th to August 8th, covering a variety of fields such as autonomous driving and speech models.

Perch 2.0, jointly launched by Google DeepMind and Google Research, has further pushed bioacoustic research to new heights.

OpenAI has officially open-sourced gpt-oss-120b and gpt-oss-20b. The former is designed for complex reasoning and knowledge-intensive scenarios, while the latter is more suitable for low-latency, local or professional vertical fields.

GPT-5 is finally here! It has significantly improved in three major scenarios: programming, creative writing, and health.