paper-with-me

홈 › Papers

Smirk: An Atomically Complete Tokenizer for Molecular Foundation Models

2024-09-19 · Alexius Wadell, Anoushka Bhutani, Venkatasubramanian Viswanathan

Text-based foundation models have become an important part of scientific discovery, with molecular foundation models accelerating advancements in molecular design and materials science. However, existing models are constrained by closed-vocabulary tokenizers which capture only a fraction of molecular space. In this work, we systematically evaluate thirty tokenizers, including 19 chemistry-specific ones, for their coverage of the SMILES molecular representation language, revealing significant gaps. To assess the impact of tokenizer choice, we introduce n-gram language models as a low-cost proxy and validate their effectiveness by training and fine-tuning 18 RoBERTa-style encoders for molecular property prediction. To overcome the limitations of existing tokenizers, we propose two new tokenizers -- Smirk and Smirk-GPE -- with full coverage of the OpenSMILES specification. Our results highlight the need for open-vocabulary modeling and chemically diverse benchmarks in cheminformatics. The proposed tokenizer framework systematically integrates nuclear, electronic, and geometric degrees of freedom; this facilitates applications in pharmacology, agriculture, biology, and energy storage.

📄 PDF Abstract BibTeX arXiv:2409.15370

Code (0)

등록된 구현이 없습니다.

Tasks

DiversityMolecular Property Predictionmolecular representationProperty Predictionscientific discovery

Methods 이 논문이 사용한 방법론

SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…

Similar Papers 제목 키워드 기반

Foundation Models for Discovery and Exploration in Chemical Space

2025-10-20 · Alexius Wadell, Anoushka Bhutani, Victor Azumah, Austin R. Ellis-Mohr 외 arxiv

Accurate prediction of atomistic, thermodynamic, and kinetic properties from molecular structures underpins materials innovation. Existing computational and experimental approaches lack the scalability required to naviga…

Ergo, SMIRK is Safe: A Safety Case for a Machine Learning Component in a Pedestrian Automatic Emergency Brake System

2022-04-16 · Markus Borg, Jens Henriksson, Kasper Socha, Olof Lennartsson 외

Integration of Machine Learning (ML) components in critical applications introduces novel challenges for software certification and verification. New safety standards and technical guidelines are under development to sup…

BIG-bench Machine Learning

Uni-Mol3: A Multi-Molecular Foundation Model for Advancing Organic Reaction Modeling

2025-07-30 · Lirong Wu, Junjie Wang, Zhifeng Gao, Xiaohong Ji 외 arxiv

Organic reaction, the foundation of modern chemical industry, is crucial for new material development and drug discovery. However, deciphering reaction mechanisms and modeling multi-molecular relationships remain formida…

Representation LearningDrug Discovery

PCICF: A Pedestrian Crossing Identification and Classification Framework

2025-09-29 · Junyi Gu, Beatriz Cabrero-Daniel, Ali Nouri, Lydia Armini 외 arxiv

We have recently observed the commercial roll-out of robotaxis in various countries. They are deployed within an operational design domain (ODD) on specific routes and environmental conditions, and are subject to continu…

NeuroMosaic: Anatomically Grounded Multimodal Large Language Modeling for Molecularly Aware Glioma Reasoning from 3D MRI and Clinical Narratives

2026-08-04 · Yantong Liu, Zheyu Zhang, Runpeng Liu, Mu Xitang 외 arxiv

Multimodal medical large language models remain structurally weak for neuro-oncology because volumetric evidence is compressed into generic visual tokens and diagnostic conclusions often lack an auditable link to MRI reg…