# Deep Molecular Dreaming: Inverse machine learning for de-novo molecular design and interpretability with surjective representations

arXiv: https://arxiv.org/abs/2012.09712 · HTML: https://mlsanity.com/paper/2012.09712/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Dec 2020.

## Built on by (most cited first, 20 of 20)

- [SELFIES and the future of molecular string representations](https://mlsanity.com/paper/2204.00056/) (2022)
- [On scientific understanding with artificial intelligence](https://mlsanity.com/paper/2204.01467/) (2022)
- [Regression Transformer: Concurrent sequence regression and generation for molecular language modeling](https://mlsanity.com/paper/2202.01338/) (2022)
- [Reinforced Genetic Algorithm for Structure-based Drug Design](https://mlsanity.com/paper/2211.16508/) (2022)
- [LIMO: Latent Inceptionism for Targeted Molecule Generation](https://mlsanity.com/paper/2206.09010/) (2022)
- [Searching for High-Value Molecules Using Reinforcement Learning and Transformers](https://mlsanity.com/paper/2310.02902/) (2023)
- [Assigning Confidence to Molecular Property Prediction](https://mlsanity.com/paper/2102.11439/) (2021)
- [Recent advances in the Self-Referencing Embedding Strings (SELFIES) library](https://mlsanity.com/paper/2302.03620/) (2023)
- [Structure-based Drug Design Benchmark: Do 3D Methods Really Dominate?](https://mlsanity.com/paper/2406.03403/) (2024)
- [Explainability Techniques for Chemical Language Models](https://mlsanity.com/paper/2305.16192/) (2023)
- [Deep Quantum Graph Dreaming: Deciphering Neural Network Insights into Quantum Experiments](https://mlsanity.com/paper/2309.07056/) (2023)
- [Quantum Deep Dreaming: A Novel Approach for Quantum Circuit Design](https://mlsanity.com/paper/2211.04343/) (2022)
- [Molecular De Novo Design through Transformer-based Reinforcement Learning](https://mlsanity.com/paper/2310.05365/) (2023)
- [Beyond Affinity: A Benchmark of 1D, 2D, and 3D Methods Reveals Critical Trade-offs in Structure-Based Drug Design](https://mlsanity.com/paper/2601.14283/) (2026)
- [An In-depth Summary of Recent Artificial Intelligence Applications in Drug Design](https://mlsanity.com/paper/2110.05478/) (2021)
- [Will More Expressive Graph Neural Networks do Better on Generative Tasks?](https://mlsanity.com/paper/2308.11978/) (2023)
- [Using GNN property predictors as molecule generators](https://mlsanity.com/paper/2406.03278/) (2024)
- [Transferable Graph Neural Fingerprint Models for Quick Response to Future Bio-Threats](https://mlsanity.com/paper/2308.01921/) (2023)
- [Quantum-inspired Reinforcement Learning for Synthesizable Drug Design](https://mlsanity.com/paper/2409.09183/) (2024)
- [Neural surrogates for designing gravitational wave detectors](https://mlsanity.com/paper/2511.19364/) (2025)

## Abstract (arXiv)

Computer-based de-novo design of functional molecules is one of the most prominent challenges in cheminformatics today. As a result, generative and evolutionary inverse designs from the field of artificial intelligence have emerged at a rapid pace, with aims to optimize molecules for a particular chemical property. These models 'indirectly' explore the chemical space; by learning latent spaces, policies, distributions or by applying mutations on populations of molecules. However, the recent development of the SELFIES string representation of molecules, a surjective alternative to SMILES, have made possible other potential techniques. Based on SELFIES, we therefore propose PASITHEA, a direct gradient-based molecule optimization that applies inceptionism techniques from computer vision. PASITHEA exploits the use of gradients by directly reversing the learning process of a neural network, which is trained to predict real-valued chemical properties. Effectively, this forms an inverse regression model, which is capable of generating molecular variants optimized for a certain property. Although our results are preliminary, we observe a shift in distribution of a chosen property during inverse-training, a clear indication of PASITHEA's viability. A striking property of inceptionism is that we can directly probe the model's understanding of the chemical space it was trained on. We expect that extending PASITHEA to larger datasets, molecules and more complex properties will lead to advances in the design of new functional molecules as well as the interpretation and explanation of machine learning models.