# RxnNano:Training Compact LLMs for Chemical Reaction and Retrosynthesis Prediction via Hierarchical Curriculum Learning

arXiv: https://arxiv.org/abs/2603.02215 · HTML: https://mlsanity.com/paper/2603.02215/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 54 references that are in the map.
- First posted on arXiv 10 Feb 2026.

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

- [Galactica: A Large Language Model for Science](https://mlsanity.com/paper/2211.09085/) (2022)
- [LlaSMol: Advancing Large Language Models for Chemistry with a Large-Scale, Comprehensive, High-Quality Instruction Tuning Dataset](https://mlsanity.com/paper/2402.09391/) (2024)
- [Retrosynthesis Prediction with Conditional Graph Logic Network](https://mlsanity.com/paper/2001.01408/) (2020)
- [Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network](https://mlsanity.com/paper/1709.04555/) (2017)
- [A Review of Large Language Models and Autonomous Agents in Chemistry](https://mlsanity.com/paper/2407.01603/) (2024)
- [Molecule Edit Graph Attention Network: Modeling Chemical Reactions as Sequences of Graph Edits](https://mlsanity.com/paper/2006.15426/) (2020)
- [Retrosynthetic reaction prediction using neural sequence-to-sequence models](https://mlsanity.com/paper/1706.01643/) (2017)
- [A Graph to Graphs Framework for Retrosynthesis Prediction](https://mlsanity.com/paper/2003.12725/) (2020)
- [RetroXpert: Decompose Retrosynthesis Prediction like a Chemist](https://mlsanity.com/paper/2011.02893/) (2020)
- [Permutation invariant graph-to-sequence model for template-free retrosynthesis and reaction prediction](https://mlsanity.com/paper/2110.09681/) (2021)
- [Learning Graph Models for Retrosynthesis Prediction](https://mlsanity.com/paper/2006.07038/) (2020)
- ["Found in Translation": Predicting Outcomes of Complex Organic Chemistry Reactions using Neural Sequence-to-Sequence Models](https://mlsanity.com/paper/1711.04810/) (2017)
- [Developing ChemDFM as a large language foundation model for chemistry](https://mlsanity.com/paper/2401.14818/) (2024)
- [Root-aligned SMILES: A Tight Representation for Chemical Reaction Prediction](https://mlsanity.com/paper/2203.11444/) (2022)
- [Re-evaluating Retrosynthesis Algorithms with Syntheseus](https://mlsanity.com/paper/2310.19796/) (2023)
- [RetroGraph: Retrosynthetic Planning with Graph Search](https://mlsanity.com/paper/2206.11477/) (2022)
- [ChemDFM-X: Towards Large Multimodal Model for Chemistry](https://mlsanity.com/paper/2409.13194/) (2024)
- [MechRetro is a chemical-mechanism-driven graph learning framework for interpretable retrosynthesis prediction and pathway planning](https://mlsanity.com/paper/2210.02630/) (2022)
- [Enhancing Chemical Reaction and Retrosynthesis Prediction with Large Language Model and Dual-task Learning](https://mlsanity.com/paper/2505.02639/) (2025)
- [UAlign: Pushing the Limit of Template-free Retrosynthesis Prediction with Unsupervised SMILES Alignment](https://mlsanity.com/paper/2404.00044/) (2024)

## Abstract (arXiv)

Chemical reaction prediction is pivotal for accelerating drug discovery and synthesis planning. Despite advances in data-driven models, current approaches are hindered by an overemphasis on parameter and dataset scaling. Some methods coupled with evaluation techniques that bypass fundamental challenges in reaction representation and fail to capture deep chemical intuition like reaction common sense and topological atom mapping logic. We argue that the core challenge lies in instilling these knowledge into the models. To this end, we propose a unified framework that prioritizes chemical understanding over scale through three key innovations: (1) a Latent Chemical Consistency objective that models reactions as movements on a continuous chemical manifold, ensuring reversible and physically plausible transformations; (2) a Hierarchical Cognitive Curriculum that trains the model through progressive stages, from syntax mastery to semantic reasoning, building robust chemical intuition; (3) Atom-Map Permutation Invariance (AMPI), which force the model to learn invariant relational topology and balance multi-task learning. (4)and structured plan-based reasoning to improve the performance of the LLMs. Our compact 0.5B-parameter model, RxnNano significantly outperforms fine-tuned LLMs ten times larger (>7B) and all the domain baselines, achieving a 23.5% Top-1 accuracy improvement on rigorous benchmarks without test-time augmentation. https://github.com/rlisml/RxnNano.