Research Tools: code, datasets, & models

Deployment of run history services in a private Azure subscription. Help research projects store experimentation data in a discoverable way.

LIDA is a library for generating grammar-agnostic (will work with any programming language and visualization libraries e.g. matplotlib, seaborn, altair, d3 etc) visualizations and infographics. LIDA comprises of 4 modules – A SUMMARIZER that converts…

Machnet is a user-level network stack deployed as a container. It provides applications with a fast, low-latency, reliable, message-based transport for communication. A micro-kernel insprided user-level network stack, that aims on performance and usability. It…

Resource Leak Checker (RLC#) for C# code using CodeQL—RLC# is a lightweight and modular resource leak checker for C# code. It is inspired by Checker Framework’s resource leak checker (RLC) for Java. RLC# is developed…

Bridging Discrete and Backpropagation: Straight-Through and Beyond—Guided by our findings, we propose a novel method called ReinMax, which integrates Heun’s Method, a second-order numerical method for solving ODEs, to approximate the gradient. Our method, ReinMax,…

Benchmarking Spatial Relationships in Text-to-Image Generation—Spatial understanding is a fundamental aspect of computer vision and integral for human-level reasoning about images, making it an important component for grounded language understanding. While recent large-scale text-to-image synthesis…

This repository contains code and data for “Accurate Measures of Vaccination and Concerns of Vaccine Holdouts from Web Search Logs” (2023) by Serina Chang, Adam Fourney, and Eric Horvitz.

CO-BED: Information-Theoretic Contextual Optimization via Bayesian Experimental Design. We formalize the problem of contextual optimization through the lens of Bayesian experimental design and propose CO-BED—a general, model-agnostic framework for designing contextual experiments using information-theoretic principles.

We present a dataset of distributed energy resources (DERs) for the contiguous U.S. using only publicly available data. The primary focus of the dataset is on distribution-level utility-scale and distributed solar and storage, given their…

We view Large Language Models as stochastic language layers in a network, where the learnable parameters are the natural language prompts at each layer. We stack two such layers, feeding the output of one layer…