A curated list of awesome AI tools, libraries, papers, datasets, and frameworks that accelerate scientific discovery — from physics and chemistry to biology, materials, and beyond.

Democratizing AI scientists with ToolUniverse

A Collection of Awesome Large Weather Models (LWMs) | AI for Earth (AI4Earth) | AI for Science (AI4Science)

Scientific machine learning (SciML) benchmarks, AI for science, and (differential) equation solvers. Covers Julia, Python (PyTorch, Jax), MATLAB, R

❓Curie: Automated and Rigorous Scientific Experimentation with AI Agents

[ICLR 2024] EquiformerV2: Improved Equivariant Transformer for Scaling to Higher-Degree Representations

The official implementation of 3D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Prediction (ICLR 2023)

[ICLR 2024] Mol-Instructions: A Large-Scale Biomolecular Instruction Dataset for Large Language Models

[ICLR 2023 Spotlight] Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs

[updating] Chinese Medical Dataset 致力于详细整理所有现有中文医学数据集，包括详细的数据汇总、数据示例、下载链接等。

Automated Hypothesis Testing with Agentic Sequential Falsifications

[NeurIPS'24] Generalizing Weather Forecast to Fine-grained Temporal Scales via Physics-AI Hybrid Modeling

[Briefings In Bioinformatics] DrugAssist: A Large Language Model for Molecule Optimization

Learning the language of protein-protein interactions

Generative Pre-trained Graph Eulerian Transformer [ICML2025]

AI for Science

Official PyTorch implementation of PSE/PSRN: Fast and efficient symbolic expression discovery through parallelized symbolic enumeration. Evaluates millions of expressions simultaneously on GPU with automated subtree reuse.

PaddleMaterials is a data-mechanism dual-driven, foundation model development and deployment, end to end toolkit based on PaddlePaddle deep learning framework for materials science and engineering.

AI for remote sensing, remote sense, object detection, oriented object detection, computer vision, cv

HypotheSAEs: hypothesizing interpretable relationships in text datasets using sparse autoencoders. https://arxiv.org/abs/2502.04382

Generating and validating natural-language explanations for the brain.

PaddleCFD is a deep learning toolkit for surrogate modeling, equation discovery, shape optimization and flow-control strategy discovery in the field of fluid mechanics.

SC23 Deep Learning at Scale Tutorial Material

Official implementation of Neural Lithography (SIGGRAPH Asia 2023)

Awesome AI for chemistry papers

[Nature Machine Intelligence] ImmunoStruct enables multimodal deep learning for immunogenicity prediction

HealthFlow: A Self-Evolving AI Agent with Meta Planning for Autonomous Healthcare Research

Dynamically constructs and adapts an LLM-generated taxonomy to a given corpus across multiple dimensions.

A Multi-Modal AI Copilot for Single-Cell Analysis with Instruction Following

Terminal-Bench-Science: Evaluating AI Agents on Complex Real-World Scientific Workflows in the Terminal