The paper reports a large-scale, compute-controlled study of Chemical Language Models (CLMs) involving over 30,000 experiments across different molecular representations, tokenizations, model sizes, datasets, and architectures. It finds clear scaling trends in pretraining loss but shows that these improvements do not translate into proportional gains in goal-directed molecular design, with chemical syntax saturating early while semantic properties develop more slowly. The authors release a new suite of models, NovoMolGen, that achieves state-of-the-art results in drug discovery tasks, highlighting a disconnect between representation learning and downstream design and calling for new pretraining paradigms that target chemical semantics.
By Roshan Balaji, Kamran Chitsaz, Quentin Fournier, Nirav Pravinbhai Bhatt, Sarath Chandar
arXiv:2602. 02320v4 Announce Type: replace-cross Abstract: Molecular function is largely determined by structure.
By Feiyang Cai, Guijuan He, Yi Hu, Jingjing Wang, Joshua Luo, Tianyu Zhu, Srikanth Pilla, Gang Li, Ling Liu, Feng Luo
arXiv:2606. 03057v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used for molecular tasks, but it remains unclear which molecular representation to use.
By Arun Raja, Garrett M. Morris, Kian Ming A. Chai
arXiv:2606. 05693v1 Announce Type: new Abstract: Large language models (LLMs) have shown promise for molecular property prediction, but their ability to reason over chemical structures remains limited, as molecular representations such as SMILES differ substantially from the natural language on which LLMs are primarily trained.
By Joey Chan, Wonbin Kweon, Ashley Shin, Niharika Bhattacharjee, Pengcheng Jiang, Yue Guo, Jiawei Han
arXiv:2608. 03855v1 Announce Type: new Abstract: Transformer models have revolutionized natural language processing (NLP), and text-based molecular representations like SMILES have successfully extended these architectures to chemistry.
By David Ming Segura, Jeremy Goumaz, Joshua W. Sin, Bojana Rankovi\'c, Philippe Schwaller
MolSC is a new dataset of 181,000 substituent-level examples that captures how attaching specific substituents to molecular scaffolds changes properties such as bioactivity and physicochemical descriptors. The authors also provide MolSC-Bench, a held‑out benchmark of 1,541 examples that are disjoint from MolSC at scaffold, substituent, and molecule levels. Experiments show that training molecular large language models on MolSC markedly improves their ability to predict substituent contributions, outperforming existing models on a range of downstream chemistry tasks.
By Hyuntae Park, Sooyeon Kim, Jiwon Park, SangKeun Lee
arXiv:2605. 16823v2 Announce Type: replace Abstract: Large language models succeed by combining large-scale pretraining with meaningful discrete tokens.
By Takayuki Kimura
arXiv:2604. 06336v2 Announce Type: replace-cross Abstract: Fragment-level representations provide a natural way to capture recurring molecular substructures and reuse their learned representations across molecules.
By Yi Yang, Ovidiu Daescu
arXiv:2606. 18703v1 Announce Type: new Abstract: Pretrained biological language models expose per-token probability distributions through masked-token prediction, providing the likelihood interface central to sequence design, variant scoring, and mechanistic interpretation.
By Yanjun Shao, Yundi Chen, Yashvi Patel, Aurelien Pelissier, Mar\'ia Rodr\'iguez Mart\'inez
arXiv:2511. 19264v2 Announce Type: replace-cross Abstract: Generative Flow Networks (GFlowNets) construct molecules through sequential decisions, but their internal policies remain opaque, limiting adoption in drug discovery, where chemists need interpretable rationales for proposed structures.
By Amirtha Varshini A S, Duminda S. Ranasinghe, Hok Hei Tam
arXiv:2606. 12113v1 Announce Type: cross Abstract: Transformer-based language models for SMILES strings suffer from a locality gap: standard character-level tokenization fragments chemically meaningful motifs, forcing models to repeatedly learn local syntax at the expense of long-range dependencies.
By Xinni Zhang, Zijing Liu, He Cao, Yu Li, Irwin King
arXiv:2607. 24848v1 Announce Type: cross Abstract: Pretrained molecular encoders are commonly evaluated through downstream prediction, but predictive accuracy alone does not establish that a learned representation captures reproducible scientific structure, adds information beyond strong conventional baselines, or transfers out of distribution.
By Kai Lun Huang (California State University, Fullerton), Wei Chieh Sun (University of Washington)