README.md

Language Model Learns to Mislead Humans via RLHF

This repository contains the RLHF'ed code generation model in our paper: https://arxiv.org/pdf/2409.12822.

It's initialized based on deepseek-coder-7B.