# Rethinking Reinforcement Learning based Logic Synthesis

arXiv: https://arxiv.org/abs/2205.07614 · HTML: https://mlsanity.com/paper/2205.07614/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 3 of its 15 references that are in the map.
- First posted on arXiv 16 May 2022.

## Builds on (most cited first, 3 of 3)

- [DRiLLS: Deep Reinforcement Learning for Logic Synthesis](https://mlsanity.com/paper/1911.04021/) (2019)
- [Developing Synthesis Flows Without Human Knowledge](https://mlsanity.com/paper/1804.05714/) (2018)
- [BOiLS: Bayesian Optimisation for Logic Synthesis](https://mlsanity.com/paper/2111.06178/) (2021)

## Built on by (most cited first, 2 of 2)

- [PUZZLES: A Benchmark for Neural Algorithmic Reasoning](https://mlsanity.com/paper/2407.00401/) (2024)
- [Adaptive Reconvergence-driven AIG Rewriting via Strategy Learning](https://mlsanity.com/paper/2312.14536/) (2023)

## Abstract (arXiv)

Recently, reinforcement learning has been used to address logic synthesis by formulating the operator sequence optimization problem as a Markov decision process. However, through extensive experiments, we find out that the learned policy makes decisions independent from the circuit features (i.e., states) and yields an operator sequence that is permutation invariant to some extent in terms of operators. Based on these findings, we develop a new RL-based method that can automatically recognize critical operators and generate common operator sequences generalizable to unseen circuits. Our algorithm is verified on both the EPFL benchmark, a private dataset and a circuit at industrial scale. Experimental results demonstrate that it achieves a good balance among delay, area and runtime, and is practical for industrial usage.