ChemBench introduces an automated benchmark of 2,700+ chemistry questions to evaluate LLMs against human expert chemists, revealing that frontier models outperform domain experts on average while struggling with basic tasks and confidence calibration.
ChemEval is a four-level, 62-task benchmark for evaluating LLMs across chemical knowledge, literature understanding, molecular reasoning, and scientific deduction, revealing that general LLMs excel at comprehension while chemistry-specific models perform better on domain tasks.
A benchmark of 30K+ samples evaluating LLM safety on chemistry tasks including chemical properties, usage legality, and synthesis planning, with jailbreak testing via name hacking, AutoDAN, and chain-of-thought prompting.
ChemDFM-R is a 14B-parameter chemical reasoning model that integrates a 101B-token dataset of atomized chemical knowledge. Using a mix-sourced distillation strategy and domain-specific reinforcement learning, it outperforms similarly sized models and DeepSeek-R1 on ChemEval.
ChemDFM-X is a multimodal chemical foundation model that integrates five non-text modalities (2D graphs, 3D conformations, images, MS2 spectra, IR spectra) into a single LLM decoder. It overcomes data scarcity by generating a 7.6M instruction-tuning dataset through approximate calculations and model predictions, establishing strong baseline performance across multiple modalities.
InstructMol integrates a pre-trained molecular graph encoder (MoleculeSTM) with a Vicuna-7B LLM using a linear projector. It employs a two-stage training process (alignment pre-training followed by task-specific instruction tuning with LoRA) to excel at property prediction, description generation, and reaction analysis.
MERMaid leverages fine-tuned vision models and VLM reasoning to mine chemical reaction data directly from PDF figures and tables. By handling context inference and coreference resolution, it builds high-fidelity knowledge graphs with 87% end-to-end accuracy.
This paper presents a multimodal search system that facilitates passage-level retrieval of chemical reactions and molecular structures by linking diagrams, text, and reaction records extracted from scientific PDFs.
A 2025 AAAI paper introducing ChemVLM, a domain-specific multimodal LLM (26B parameters). It achieves state-of-the-art performance on chemical OCR, reasoning benchmarks, and molecular understanding tasks by combining vision and language models trained on curated chemistry data.
ZINC-22 is a multi-billion-scale public database containing over 37 billion make-on-demand molecules. It utilizes distributed infrastructure and specialized search algorithms to support modern ultra-large virtual screening campaigns.
MARCEL provides a comprehensive benchmark for molecular representation learning with 722K+ conformers across four diverse subsets (Drugs-75K, Kraken, EE, BDE), enabling evaluation of conformer ensemble methods for property prediction in drug discovery and catalysis.
GEOM contains 450k+ molecules with 37M+ conformations, featuring energy annotations from semi-empirical (GFN2-xTB) and DFT methods for property prediction and molecular generation research.