Dataset Source Code
Sui Generis
Source code for the paper: Echoes in AI: Quantifying Lack of Plot Diversity in LLM Outputs
Research Tools: code, datasets, & models
Discover an index of datasets, SDKs, APIs and open-source tools developed by Microsoft researchers and shared with the global academic community below. These experimental technologies—available through Azure AI Foundry Labs (opens in new tab)—offer a glimpse into the future of AI innovation.
Tool
TRELLIS
TRELLIS is a large 3D asset generation model that creates high-quality 3D assets from simple text or image inputs. Using a unified latent space (SLAT), it delivers detailed, textured 3D models in formats like meshes,…
Dataset Source Code
ChatBench
ChatBench Interactive Benchmark Simulator enables automated, realistic evaluation of AI models through simulated user-AI conversations. We release fine-tuned user simulators (model weights) and supporting infrastructure, allowing the community to assess models in interactive, user-in-the-loop scenarios.
Dataset Source Code
SimulatorArena
This repository contains the code and data for SimulatorArena, a framework that enables: (1) benchmarking AI assistants through multi-turn conversations with user simulators, and (2) evaluating the reliability of user simulators as proxies for human…
Tool
MatterSim
MatterSim is a deep learning model for accurate and efficient materials simulation and property prediction over a broad range of elements, temperatures and pressures to enable in silico materials design.
Dataset Source Code
BAPO – Bounded Attention Prefix Oracle
This repository contains all scripts for re-producing the results of our paper “Lost in Transmission: When and Why LLMs Fail to Reason Globally”.
Tool
ExACT
ExACT is an approach for teaching AI agents to explore more effectively, enabling them to intelligently navigate their environments, gather valuable information, evaluate options, and identify optimal decision-making and planning strategies.
Tool
Microsoft Research Accurate Chemistry Collection (MSR-ACC)
The Skala functional will enable more accurate, scalable predictions in computational chemistry. It starts with the largest high-accuracy dataset ever built for training deep-learning-based density functional theory (DFT) models. This dataset underpins Skala—coming soon to…