# microsoft/bingo

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/microsoft-bingo).**

0 stars · 0 forks

## Links

- GitHub: https://github.com/microsoft/Bingo
- awesome-repositories: https://awesome-repositories.com/repository/microsoft-bingo.md

## Description

We propose Bingo, an RL framework that enhances efficient reasoning in LLMs through advanced length-based reward design. It incorporates two key mechanisms: a significance-aware length reward that gradually guides the model to reduce only insignificant tokens, and a dynamic length reward that…

## Tags

### Part of an Awesome List

- [Regularization Objectives](https://awesome-repositories.com/f/awesome-lists/ai/regularization-objectives.md) — Dynamic and significance-based reinforcement learning for efficiency.