# Look the Other Way: Designing 'Positive' Molecules with Negative Data via Task Arithmetic

arXiv: https://arxiv.org/abs/2507.17876 · HTML: https://mlsanity.com/paper/2507.17876/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 77 references that are in the map.
- First posted on arXiv 23 Jul 2025.

## Builds on (most cited first, 24 of 24)

- [Learning without Forgetting](https://mlsanity.com/paper/1606.09282/) (2016)
- [Gradient Episodic Memory for Continual Learning](https://mlsanity.com/paper/1706.08840/) (2017)
- [Editing Models with Task Arithmetic](https://mlsanity.com/paper/2212.04089/) (2022)
- [Automatic chemical design using a data-driven continuous representation of molecules](https://mlsanity.com/paper/1610.02415/) (2016)
- [An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning](https://mlsanity.com/paper/2308.08747/) (2023)
- [GraphVAE: Towards Generation of Small Graphs Using Variational Autoencoders](https://mlsanity.com/paper/1802.03480/) (2018)
- [GuacaMol: Benchmarking Models for De Novo Molecular Design](https://mlsanity.com/paper/1811.09621/) (2018)
- [Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models](https://mlsanity.com/paper/1811.12823/) (2018)
- [Molecular De Novo Design through Deep Reinforcement Learning](https://mlsanity.com/paper/1704.07555/) (2017)
- [Constrained Graph Variational Autoencoders for Molecule Design](https://mlsanity.com/paper/1805.09076/) (2018)
- [SMILES Enumeration as Data Augmentation for Neural Network Modeling of Molecules](https://mlsanity.com/paper/1703.07076/) (2017)
- [On the generalization of language models from in-context learning and finetuning: a controlled study](https://mlsanity.com/paper/2505.00661/) (2025)
- [DOCKSTRING: easy molecular docking yields better benchmarks for ligand design](https://mlsanity.com/paper/2110.15486/) (2021)
- [A Survey on Graph Diffusion Models: Generative AI in Science for Molecule, Protein and Material](https://mlsanity.com/paper/2304.01565/) (2023)
- [How Much Space Has Been Explored? Measuring the Chemical Space Covered by Databases and Machine-Generated Molecules](https://mlsanity.com/paper/2112.12542/) (2021)
- [Revisiting Weight Averaging for Model Merging](https://mlsanity.com/paper/2412.12153/) (2024)
- [Generative chemistry: drug discovery with deep learning generative models](https://mlsanity.com/paper/2008.09000/) (2020)
- [Task Vectors in In-Context Learning: Emergence, Formation, and Benefit](https://mlsanity.com/paper/2501.09240/) (2025)
- [A Few More Examples May Be Worth Billions of Parameters](https://mlsanity.com/paper/2110.04374/) (2021)
- [Diffusion Models for Molecules: A Survey of Methods and Tasks](https://mlsanity.com/paper/2502.09511/) (2025)
- [How Much Data is Enough Data? Fine-Tuning Large Language Models for In-House Translation: Performance Evaluation Across Multiple Dataset Sizes](https://mlsanity.com/paper/2409.03454/) (2024)
- [Beyond Task Vectors: Selective Task Arithmetic Based on Importance Metrics](https://mlsanity.com/paper/2411.16139/) (2024)
- [Synergy Effect between Convolutional Neural Networks and the Multiplicity of SMILES for Improvement of Molecular Prediction](https://mlsanity.com/paper/1812.04439/) (2018)
- [Investigating Task Arithmetic for Zero-Shot Information Retrieval](https://mlsanity.com/paper/2505.00649/) (2025)

## Abstract (arXiv)

The scarcity of molecules with desirable properties (i.e., `positive' molecules) is an inherent bottleneck for generative molecule design. To sidestep such obstacle, here we propose molecular task arithmetic: training a model on diverse and abundant negative examples to learn 'property directions' - without accessing any positively labeled data - and moving models in the opposite property directions to generate positive molecules. When analyzed on 33 design experiments with distinct molecular entities (small molecules, proteins), model architectures, and scales, molecular task arithmetic generated more diverse and successful designs than models trained on positive molecules in general. Moreover, we employed molecular task arithmetic in dual-objective and few-shot design tasks. We find that molecular task arithmetic can consistently increase the diversity of designs while maintaining desirable complex design properties, such as good docking scores to a protein. With its simplicity, data efficiency, and performance, molecular task arithmetic bears the potential to become the de facto transfer learning strategy for de novo molecule design.