🥉 useful helpers for react-three-fiber

An extensible benchmark for evaluating large language models on planning

The official code release for Don't Trust: Verify -- Grounding LLM Quantitative Reasoning with Autoformalization

Intuitive, type-safe expression quotations for Lean 4.

The math library of Lean 4

Making Google Deepmind's AlphaGeometry accessible to the Masses

Java Geometry Expert

Inference Combinators in JAX

Synthetic question-answering dataset to formally analyze the chain-of-thought output of large language models on a reasoning task.

An extensible benchmark for evaluating large language models on planning

AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.

Materials for ConceptARC paper

RAVEN: A Dataset for Relational and Analogical Visual rEasoNing

An AI agent that solves Raven's Progressive Matrices

Official code for CVPR 2022 paper "Rethinking Visual Geo-localization for Large-Scale Applications"

Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them

A multithread Pushshift.io API Wrapper for reddit.com comment and submission searches.

Drive a browser with GPT-3

WebChatGPT: A browser extension that augments your ChatGPT prompts with web results.

Failure archive for ChatGPT and similar models

🦜🔗 Build context-aware reasoning applications

Code for Cicero, an AI agent that plays the game of Diplomacy with open-domain natural language negotiation.