-
Carnegie Mellon University
- Pittsburgh
Highlights
- Pro
Stars
ImProver: Agent-Based Automated Proof Optimization
Evaluation for miniCodeProps: a Minimal Benchmark for Proving Code Properties
Neural theorem proving evaluation via the Lean REPL
Neural theorem proving toolkit: data extraction tools for Lean 4
Evaluate your LLM's response with Prometheus and GPT4 💯
Neural theorem proving tutorial, version II
Easy-to-Hard Generalization: Scalable Alignment Beyond Human Supervision
Simple and efficient pytorch-native transformer training and inference (batched)
Tutorial on neural theorem proving
A high-throughput and memory-efficient inference and serving engine for LLMs
zhangir-azerbayev / repl
Forked from leanprover-community/replA simple REPL for Lean 4, returning information about errors and sorries.
llmstep: [L]LM proofstep suggestions in Lean 4.
A simple REPL for Lean 4, returning information about errors and sorries.
An environment for learning formal mathematical reasoning from scratch
Benchmark for undergraduate-level formal mathematics
An updated version of miniF2F with lots of fixes and informal statements / solutions.
PyTorch + HuggingFace code for RetoMaton: "Neuro-Symbolic Language Modeling with Automaton-augmented Retrieval" (ICML 2022), including an implementation of kNN-LM and kNN-MT
An open bibliography of machine learning for formal proof papers
https://albertqjiang.github.io/Portal-to-ISAbelle/
NaturalProver: Grounded Mathematical Proof Generation with Language Models