# BF++: a language for general-purpose program synthesis

arXiv: https://arxiv.org/abs/2101.09571 · HTML: https://mlsanity.com/paper/2101.09571/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 67 references that are in the map.
- First posted on arXiv 23 Jan 2021.

## Builds on (most cited first, 20 of 20)

- [OpenAI Gym](https://mlsanity.com/paper/1606.01540/) (2016)
- [Dota 2 with Large Scale Deep Reinforcement Learning](https://mlsanity.com/paper/1912.06680/) (2019)
- [A Brief Survey of Deep Reinforcement Learning](https://mlsanity.com/paper/1708.05866/) (2017)
- [Memory Networks](https://mlsanity.com/paper/1410.3916/) (2014)
- [Deep Reinforcement Learning: An Overview](https://mlsanity.com/paper/1701.07274/) (2017)
- [Reinforcement Learning in Healthcare: A Survey](https://mlsanity.com/paper/1908.08796/) (2019)
- [Neural-Symbolic Learning and Reasoning: A Survey and Interpretation](https://mlsanity.com/paper/1711.03902/) (2017)
- [Programmatically Interpretable Reinforcement Learning](https://mlsanity.com/paper/1804.02477/) (2018)
- [Learning to Mine Aligned Code and Natural Language Pairs from Stack Overflow](https://mlsanity.com/paper/1805.08949/) (2018)
- [SQLNet: Generating Structured Queries From Natural Language Without Reinforcement Learning](https://mlsanity.com/paper/1711.04436/) (2017)
- [A Neural Model for Generating Natural Language Summaries of Program Subroutines](https://mlsanity.com/paper/1902.01954/) (2019)
- [Neural-Guided Deductive Search for Real-Time Program Synthesis from Examples](https://mlsanity.com/paper/1804.01186/) (2018)
- [Neural Random-Access Machines](https://mlsanity.com/paper/1511.06392/) (2015)
- [Neural Program Synthesis with Priority Queue Training](https://mlsanity.com/paper/1801.03526/) (2018)
- [Synthetic Datasets for Neural Program Synthesis](https://mlsanity.com/paper/1912.12345/) (2019)
- [TerpreT: A Probabilistic Programming Language for Program Induction](https://mlsanity.com/paper/1608.04428/) (2016)
- [Reinforcement Learning Neural Turing Machines - Revised](https://mlsanity.com/paper/1505.00521/) (2015)
- [Zooming for Efficient Model-Free Reinforcement Learning in Metric Spaces](https://mlsanity.com/paper/2003.04069/) (2020)
- [DAQN: Deep Auto-encoder and Q-Network](https://mlsanity.com/paper/1806.00630/) (2018)
- [Recent Advances in Neural Program Synthesis](https://mlsanity.com/paper/1802.02353/) (2018)

## Built on by (most cited first, 3 of 3)

- [Fully Autonomous Programming with Large Language Models](https://mlsanity.com/paper/2304.10423/) (2023)
- [Fully Autonomous Programming using Iterative Multi-Agent Debugging with Large Language Models](https://mlsanity.com/paper/2503.07693/) (2025)
- [BASIL: Best-Action Symbolic Interpretable Learning for Evolving Compact RL Policies](https://mlsanity.com/paper/2506.00328/) (2025)

## Abstract (arXiv)

Most state of the art decision systems based on Reinforcement Learning (RL) are data-driven black-box neural models, where it is often difficult to incorporate expert knowledge into the models or let experts review and validate the learned decision mechanisms. Knowledge-insertion and model review are important requirements in many applications involving human health and safety. One way to bridge the gap between data and knowledge driven systems is program synthesis: replacing a neural network that outputs decisions with a symbolic program generated by a neural network or by means of genetic programming. We propose a new programming language, BF++, designed specifically for automatic programming of agents in a Partially Observable Markov Decision Process (POMDP) setting and apply neural program synthesis to solve standard OpenAI Gym benchmarks.