# Hunter Heidenreich | Senior AI Research Scientist
- [About Hunter Heidenreich](https://hunterheidenreich.com/about/)

## Research
- [Where to Cut, How Deep: BPE and Unigram-LM on SMILES](https://hunterheidenreich.com/research/bpe-unigram-lm-smiles-vocabularies/)
- [GutenOCR: A Grounded Vision-Language Front-End for Documents](https://hunterheidenreich.com/research/gutenocr-grounded-vision-language-frontend/)
- [Optimizing Sequence Models for Dynamical Systems](https://hunterheidenreich.com/research/deconstructing-recurrence-attention-gating/)
- [PubMed-OCR: PMC Open Access OCR Annotations](https://hunterheidenreich.com/research/pubmed-ocr-pmc-open-access-ocr-annotations/)
- [LLMs for Insurance Document Automation](https://hunterheidenreich.com/research/page-stream-segmentation-llms/)
- [LLMs for Page Stream Segmentation](https://hunterheidenreich.com/research/llm-page-stream-segmentation/)
- [Analytical Solution to Word2Vec Softmax & Bias Probing](https://hunterheidenreich.com/research/word-company-vicinity/)
- [EigenNoise: Data-Free Word Vector Initialization](https://hunterheidenreich.com/research/eigennoise-contrastive-prior/)
- [Look, Don't Tweet: Unified Data Models for Social NLP](https://hunterheidenreich.com/research/look-dont-tweet/)
- [GPT-2 Susceptibility to Universal Adversarial Triggers](https://hunterheidenreich.com/research/gpt2-adversarial-triggers/)
- [NewsTweet Dataset: Social Media in Digital Journalism](https://hunterheidenreich.com/research/newstweet-social-media-journalism/)
- [Coordinated Social Targeting on Twitter](https://hunterheidenreich.com/research/coordinated-social-targeting/)
- [Data-Driven WordNet Construction from Wiktionary](https://hunterheidenreich.com/research/semantic-network-induction/)

## Projects
- [Kabsch-Horn Cookbook: Differentiable Alignment](https://hunterheidenreich.com/projects/kabsch-horn-cookbook/)
- [Molecular String Renderer: Chemical Visualization Library](https://hunterheidenreich.com/projects/molecular-string-renderer/)
- [Müller-Brown Potential: A PyTorch ML Testbed](https://hunterheidenreich.com/projects/muller-brown-pytorch/)
- [Modernizing Rahman''s 1964 Argon Simulation](https://hunterheidenreich.com/projects/rahman-1964-replication/)
- [Vectorized Word2Vec in Pure PyTorch](https://hunterheidenreich.com/projects/modern-word2vec/)
- [Synthetic Isomer Data Generation Pipeline](https://hunterheidenreich.com/projects/isomer-dataset-generation/)
- [Automated Adatom Diffusion Workflow](https://hunterheidenreich.com/projects/lammps-adatom-diffusion/)
- [Mini-Protein Trajectory Generation](https://hunterheidenreich.com/projects/mini-protein-trajectories/)
- [Congressional Knowledge Graph & Policy Classification](https://hunterheidenreich.com/projects/congressional-data-analysis/)
- [IQCRNN: Certified Stability for Neural Networks](https://hunterheidenreich.com/projects/iqcrnn-pytorch/)
- [PyConversations: Social Media Conversational Analysis](https://hunterheidenreich.com/projects/pyconversations-social-media-analysis/)
- [Cartesian Genetic Programming in Julia](https://hunterheidenreich.com/projects/cgp-julia/)
- [FFTW Compiler in Haskell](https://hunterheidenreich.com/projects/fftw-compiler-haskell/)
- [Term Schedule Optimizer](https://hunterheidenreich.com/projects/term-schedule-optimizer/)
- [Rubik's Cube Sonification](https://hunterheidenreich.com/projects/rubiks-cube-player/)
- [Elemental Brawl](https://hunterheidenreich.com/projects/elemental-brawl/)

## Blog Posts
- [The Reliability Trap: The Limits of 99% Accuracy](https://hunterheidenreich.com/posts/reliability-trap-document-automation/)
- [The Evolution of Page Stream Segmentation: Rules to LLMs](https://hunterheidenreich.com/posts/history-of-page-stream-segmentation/)
- [Importance Weighted Autoencoders: Beyond the Standard VAE](https://hunterheidenreich.com/posts/importance-weighted-autoencoders/)
- [What is Optical Chemical Structure Recognition (OCSR)?](https://hunterheidenreich.com/posts/what-is-ocsr/)
- [Converting SMILES and SELFIES to 2D Molecular Images](https://hunterheidenreich.com/posts/visualizing-smiles-and-selfies-strings/)
- [Exponential Random Numbers: Two Classic Algorithms](https://hunterheidenreich.com/posts/random-number-tricks/)
- [Implementing the Müller-Brown Potential in PyTorch](https://hunterheidenreich.com/posts/muller-brown-in-pytorch/)
- [Modernizing Rahman's 1964 Argon Simulation](https://hunterheidenreich.com/posts/rahman-1964-lammps-liquid-argon/)
- [GEOM Dataset: 3D Molecular Conformer Generation](https://hunterheidenreich.com/posts/geom-conformer-generation-dataset/)
- [Modern PyTorch VAEs: A Detailed Implementation Guide](https://hunterheidenreich.com/posts/modern-variational-autoencoder-in-pytorch/)
- [Sarcasm Detection with Transformers: A Cautionary Tale](https://hunterheidenreich.com/posts/sarcasm-detection-with-transformers/)
- [Hearing Molecular Shape via Coulomb Matrix Eigenvalues](https://hunterheidenreich.com/posts/alkane-constitutional-isomer-classification/)
- [Classifying Congressional Bills with Machine Learning](https://hunterheidenreich.com/posts/congressional-bill-policy-area-classification/)
- [Coulomb Matrices for Molecular Machine Learning](https://hunterheidenreich.com/posts/molecular-descriptor-coulomb-matrix/)
- [How Does Congress Actually Work? Data from 15K Bills](https://hunterheidenreich.com/posts/us-117th-congress-data-exploration/)
- [Kabsch Algorithm: NumPy, PyTorch, TensorFlow, and JAX](https://hunterheidenreich.com/posts/kabsch-algorithm/)
- [LAMMPS Tutorial: Copper and Platinum Adatom Diffusion](https://hunterheidenreich.com/posts/adatom-cu-diffusion/)
- [Generating Mini-Protein Trajectories with GROMACS](https://hunterheidenreich.com/posts/mini-proteins/)
- [5 Axes of Multi-Arm Bandit Problems: A Practical Guide](https://hunterheidenreich.com/posts/a-roadmap-to-multi-arm-bandit-algorithms/)
- [A Guide to Neuroevolution: NEAT and HyperNEAT](https://hunterheidenreich.com/posts/neuroevolution-neat-and-hyperneat/)
- [Breaking Down Machine Learning for the Average Person](https://hunterheidenreich.com/posts/breaking-down-ml-for-the-average-person/)
- [Foundations of AI: Knowledge-Based Agents and Logic](https://hunterheidenreich.com/posts/knowledge-based-agents-and-logic/)
- [QuAC: Question Answering in Context Dataset](https://hunterheidenreich.com/posts/quac-question-answering-in-context/)
- [CoQA Dataset: Advancing Conversational Question Answering](https://hunterheidenreich.com/posts/coqa-conversation-question-answering/)
- [Understanding GANs: From Fundamentals to Objective Functions](https://hunterheidenreich.com/posts/what-is-a-gan/)
- [Word Embeddings in NLP: An Introduction](https://hunterheidenreich.com/posts/intro-to-word-embeddings/)

## Notes

### Chemistry Research Notes

#### Molecular Simulation

##### Neural Network Potentials and Modern Methods
- [MB-nrg in Solution: Polyalanine in Water with CCSD(T) PEFs](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/mb-nrg-polyalanine-water/)
- [MB-nrg: CCSD(T)-Accurate Potentials for Polyalanine](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/mb-nrg-polyalanine-ccsdt/)
- [Atom-Density Representations for Machine Learning](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/atom-density-representations-ml/)
- [Conformation Autoencoder for 3D Molecules](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/conformation-autoencoder/)
- [Ewald Message Passing for Molecular Graphs](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/ewald-message-passing-molecular-graphs/)
- [PharMolixFM: Multi-Modal All-Atom Molecular Models](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/pharmolixfm-all-atom-foundation-models/)
- [MAT: Graph-Augmented Transformer for Molecules (2020)](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/molecule-attention-transformer/)
- [MOFFlow: Flow Matching for MOF Structure Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/mofflow/)
- [DenoiseVAE: Adaptive Noise for Molecular Pre-training](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/denoise-vae/)
- [Beyond Atoms: 3D Space Modeling for Molecular Pretraining](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/beyond-atoms/)
- [Dark Side of Forces: Non-Conservative ML Force Models](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/dark-side-of-forces/)
- [Efficient DFT Hamiltonian Prediction via Adaptive Sparsity](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/efficient-dft-hamiltonian-predicton-sphnet/)
- [eSEN: Smooth Interatomic Potentials (ICML Spotlight)](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/ml-potentials/learning-smooth-interatomic-potentials/)

##### Classical Potentials and Simulation Methods
- [Embedded-Atom Method User Guide: Voter's 1994 Chapter](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/embedded-atom-method-voter-1994/)
- [Embedded-Atom Method: Theory and Applications Review](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/embedded-atom-method-review-1993/)
- [Evans 1986: Thermal Conductivity of Lennard-Jones Fluid](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/evans-thermal-conductivity-1986/)
- [Second-Order Langevin Equation for Field Simulations](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/second-order-langevin-1987/)
- [Stillinger-Weber Potential for Silicon Simulation](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/stillinger-weber-1985/)
- [Correlations in the Motion of Atoms in Liquid Argon](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/correlations-motion-atoms-liquid-argon/)
- [The Müller-Brown Potential: A 2D Benchmark Surface](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/muller-brown-1979/)
- [Embedded-Atom Method: Impurities and Defects in Metals](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/embedded-atom-method/)
- [Umbrella Sampling: Monte Carlo Free-Energy Estimation](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/umbrella-sampling/)
- [Lennard-Jones on Adsorption and Diffusion on Surfaces](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/classical-methods/processes-of-adsorption/)

##### Surface Catalysis and Diffusion
- [Dynamical Corrections to TST for Surface Diffusion](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/surface-science/self-diffusion-lj-fcc111-1989/)
- [In Situ XRD of Oxidation-Reduction Oscillations on Pt/SiO2](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/surface-science/oxidation-reduction-oscillations-pt-sio2-1994/)
- [Kinetic Oscillations in CO Oxidation on Pt(100): Theory](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/surface-science/kinetic-oscillations-pt100-1985/)
- [MD Simulation of Self-Diffusion on Metal Surfaces (1994)](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/surface-science/self-diffusion-metal-surfaces-1994/)
- [Oscillatory CO Oxidation on Pt(110): Temporal Modeling](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/surface-science/oscillatory-co-oxidation-pt110-1992/)
- [Adatom Dimer Diffusion on fcc(111) Crystal Surfaces](https://hunterheidenreich.com/notes/chemistry/molecular-simulation/surface-science/diffusion-adatom-dimers-1984/)

#### Molecular Databases & Datasets
- [FDB-17: Fragment Database (10M Molecules)](https://hunterheidenreich.com/notes/chemistry/datasets/fdb-17/)
- [GDBMedChem: Drug-Like Subset of GDB-17 (10M Molecules)](https://hunterheidenreich.com/notes/chemistry/datasets/gdb-medchem/)
- [QM9: Quantum Chemistry Properties of 134k Molecules](https://hunterheidenreich.com/notes/chemistry/datasets/qm9/)
- [VQM24: 836k Molecules at DFT and Diffusion QMC](https://hunterheidenreich.com/notes/chemistry/datasets/vqm24/)
- [ZINC-22: A Multi-Billion Scale Database for Ligand Discovery](https://hunterheidenreich.com/notes/chemistry/datasets/zinc-22/)
- [MARCEL: Molecular Conformer Ensemble Learning Benchmark](https://hunterheidenreich.com/notes/chemistry/datasets/marcel/)
- [GEOM: Energy-Annotated Molecular Conformations Dataset](https://hunterheidenreich.com/notes/chemistry/datasets/geom/)
- [GDB-11: Chemical Universe Database (26.4M Molecules)](https://hunterheidenreich.com/notes/chemistry/datasets/gdb-11/)
- [GDB-13: Chemical Universe Database (970M Molecules)](https://hunterheidenreich.com/notes/chemistry/datasets/gdb-13/)
- [GDB-17: Chemical Universe Database (166.4B Molecules)](https://hunterheidenreich.com/notes/chemistry/datasets/gdb-17/)

#### Molecular Design

##### Chemical Space
- [ACSESS: Diverse Optimal Molecules in the SMU](https://hunterheidenreich.com/notes/chemistry/molecular-design/chemical-space/acsess-diverse-optimal-molecules/)
- [AllChem: Generating and Searching 10^20 Structures](https://hunterheidenreich.com/notes/chemistry/molecular-design/chemical-space/allchem-synthetically-accessible-structures/)
- [CHX8: Complete Eight-Carbon Hydrocarbon Space](https://hunterheidenreich.com/notes/chemistry/molecular-design/chemical-space/chx8-hydrocarbon-chemical-space/)
- [Molecular Complexity from the GDB Chemical Space](https://hunterheidenreich.com/notes/chemistry/molecular-design/chemical-space/gdb-molecular-complexity/)
- [Surge: Fastest Open-Source Chemical Graph Generator](https://hunterheidenreich.com/notes/chemistry/molecular-design/chemical-space/surge-chemical-graph-generator/)
- [VEHICLe: Heteroaromatic Rings of the Future](https://hunterheidenreich.com/notes/chemistry/molecular-design/chemical-space/vehicle-heteroaromatic-rings/)

##### Molecular Generation

###### Autoregressive Generation
- [LSTM Neural Network for Drug-Like Molecule Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/lstm-drug-like-molecule-generation/)
- [LMs Generate 3D Molecules from XYZ, CIF, PDB Files](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/3d-chemical-language-models-xyz-cif-pdb/)
- [S4 Structured State Space Models for De Novo Drug Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/s4-chemical-language-modeling/)
- [Back Translation for Semi-Supervised Molecule Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/back-translation-molecule-generation/)
- [RetMol: Retrieval-Based Controllable Molecule Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/retmol-retrieval-molecule-generation/)
- [MolGen: Molecular Generation with Chemical Feedback](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/molgen-molecular-generation-chemical-feedback/)
- [GP-MoLFormer: Molecular Generation via Transformers](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/gp-molformer/)
- [Chemformer: A Pre-trained Transformer for Comp Chem](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/autoregressive/chemformer/)

###### Latent-Space Generation
- [Grammar VAE: Generating Valid Molecules via CFGs](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/latent-space/grammar-variational-autoencoder/)
- [LatentGAN: Latent-Space GAN for Molecular Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/latent-space/latentgan-de-novo-molecular-generation/)
- [CogMol: Controlled Molecule Generation for COVID-19](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/latent-space/cogmol-target-specific-drug-design/)
- [PASITHEA: Gradient-Based Molecular Design via Dreaming](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/latent-space/deep-molecular-dreaming-pasithea/)
- [VAE for Automatic Chemical Design (2018 Seminal)](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/latent-space/automatic-chemical-design-vae/)
- [LIMO: Latent Inceptionism for Targeted Molecule Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/latent-space/limo-latent-inceptionism/)

###### RL-Tuned Generation
- [DrugEx v2: Pareto Multi-Objective RL for Drug Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/drugex-v2-pareto-multi-objective-rl/)
- [Memory-Assisted RL for Diverse De Novo Mol. Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/memory-assisted-rl-diverse-molecular-design/)
- [MolecularRNN: Graph-Based Molecular Generation and RL](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/molecularrnn-graph-generation-optimized-properties/)
- [ORGAN: Objective-Reinforced GANs for Molecule Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/organ-objective-reinforced-gan/)
- [REINVENT: Reinforcement Learning for Mol. Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/reinvent-deep-rl-molecular-design/)
- [Augmented Hill-Climb for RL-Based Molecule Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/augmented-hill-climb-rl-molecule-generation/)
- [Curriculum Learning for De Novo Drug Design (REINVENT)](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/curriculum-learning-molecular-design/)
- [DrugEx v3: Scaffold-Constrained Graph Transformer](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/drugex-v3-scaffold-graph-transformer/)
- [Link-INVENT: RL-Driven Molecular Linker Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/link-invent-generative-linker-design/)
- [REINVENT 4: Open-Source Generative Molecule Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/rl-tuned/reinvent4-generative-molecule-design/)

###### Search-Based Generation
- [ChemGE: Molecule Generation via Grammatical Evolution](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/search-based/chemge-grammatical-evolution-molecule-generation/)
- [Graph-Based GA and MCTS Generative Model for Molecules](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/search-based/graph-based-genetic-algorithm-chemical-space/)
- [STONED: Training-Free Molecular Design with SELFIES](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/search-based/stoned-selfies-chemical-space-exploration/)
- [Genetic Algorithms as Baselines for Molecule Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/search-based/genetic-algorithms-molecule-generation-baselines/)

###### Evaluation, Benchmarks & Surveys
- [Avoiding Failure Modes in Goal-Directed Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/avoiding-failure-modes-goal-directed-generation/)
- [Chemical Language Models for De Novo Drug Design Review](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/clms-de-novo-drug-design-review/)
- [Generative AI Survey for De Novo Molecule and Protein Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/generative-ai-drug-design-survey/)
- [Inverse Molecular Design with ML Generative Models](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/inverse-molecular-design-ml-review/)
- [Re-evaluating Sample Efficiency in Molecule Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/sample-efficiency-de-novo-generation/)
- [Review: Deep Learning for Molecular Design (2019)](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/deep-learning-molecular-design-review/)
- [RNNs vs Transformers for Molecular Generation Tasks](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/molecular-language-models-rnns-or-transformer/)
- [DOCKSTRING: Docking-Based Benchmarks for Drug Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/dockstring-docking-benchmarks-ligand-design/)
- [Failure Modes in Molecule Generation & Optimization](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/failure-modes-molecule-generation/)
- [Frechet ChemNet Distance for Molecular Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/frechet-chemnet-distance/)
- [GuacaMol: Benchmarking Models for De Novo Molecular Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/guacamol-benchmarking-de-novo-molecular-design/)
- [MolGenBench: Benchmarking Molecular Generative Models](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/molgenbench-molecular-generative-models/)
- [MolScore: Scoring and Benchmarking for Drug Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/molscore-scoring-benchmarking-framework/)
- [PMO: Benchmarking Sample-Efficient Molecular Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/pmo-sample-efficient-molecular-optimization/)
- [SPECTRA: Evaluating Generalizability of Molecular AI](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/spectra-evaluating-generalizability-molecular-ai/)
- [MolGenSurvey: Systematic Survey of ML for Molecule Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/molgensurvey-molecule-design/)
- [SMINA Docking Benchmark for De Novo Drug Design Models](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/smina-docking-benchmark/)
- [Tartarus: Realistic Inverse Molecular Design Benchmarks](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/tartarus-inverse-molecular-design/)
- [UnCorrupt SMILES: Post Hoc Correction for De Novo Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/uncorrupt-smiles/)
- [Molecular Sets (MOSES): A Generative Modeling Benchmark](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/evaluation/molecular-sets-moses/)

###### Target-Aware Generation
- [AlphaDrug: MCTS-Guided Target-Specific Drug Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/target-aware/alphadrug-protein-target-molecular-generation/)
- [BindGPT: GPT for 3D Molecular Design and Docking](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/target-aware/bindgpt-3d-molecular-design/)
- [Evolutionary Molecular Design via Deep Learning + GA](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/target-aware/evolutionary-design-deep-learning-genetic-algorithm/)
- [Lingo3DMol: Language Model for 3D Molecule Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/target-aware/lingo3dmol-3d-molecule-generation/)
- [PrefixMol: Prefix Embeddings for Drug Molecule Design](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/target-aware/prefixmol-target-chemistry-aware-generation/)
- [Protein-to-Drug Molecule Translation via Transformer](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/target-aware/transformer-protein-drug-generation/)
- [TamGen: GPT-Based Target-Aware Drug Design and Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/generation/target-aware/tamgen-target-aware-molecule-generation/)

##### Reaction Prediction
- [Data Transfer Approaches for Seq-to-Seq Retrosynthesis](https://hunterheidenreich.com/notes/chemistry/molecular-design/reaction-prediction/data-transfer-seq-to-seq-retrosynthesis/)
- [Neural Machine Translation for Reaction Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-design/reaction-prediction/nmt-organic-reaction-prediction/)
- [ReactionT5: Pre-trained T5 for Reaction Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-design/reaction-prediction/reactiont5-pretrained-limited-reaction-data/)
- [Tied Two-Way Transformers for Diverse Retrosynthesis](https://hunterheidenreich.com/notes/chemistry/molecular-design/reaction-prediction/tied-two-way-transformers-retrosynthesis/)
- [Molecular Transformer: Calibrated Reaction Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-design/reaction-prediction/molecular-transformer/)

##### Property Prediction
- [Maxsmi: SMILES Augmentation for Property Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/maxsmi-smiles-augmentation-property-prediction/)
- [MTL-BERT: Multitask BERT for Property Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/mtl-bert-multitask-smiles-enumeration/)
- [LLM-Prop: Predicting Crystal Properties from Text](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/llm-prop-crystal-property-prediction/)
- [MolPMoFiT: Inductive Transfer Learning for QSAR](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/molpmofit-transfer-learning-qsar/)
- [SMILES2Vec: Interpretable Chemical Property Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/smiles2vec-interpretable-property-prediction/)
- [Transformer-CNN: SMILES Embeddings for QSAR Modeling](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/transformer-cnn-qsar-modeling/)
- [Transformers for Molecular Property Prediction Review](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/transformers-molecular-property-prediction-review/)
- [Benchmarking Molecular Property Prediction at Scale](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/systematic-study-molecular-property-prediction/)
- [MoleculeNet: Benchmarking Molecular Machine Learning](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/moleculenet-benchmark-molecular-ml/)
- [Perplexity for Molecule Ranking and CLM Bias Detection](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/perplexity-molecule-ranking-bias-clms/)
- [ROGI-XD: Roughness of Pretrained Molecular Representations](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/rogi-xd-roughness-pretrained-representations/)
- [Language Models Learn Complex Molecular Distributions](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/lm-complex-molecular-distributions/)
- [Regression Transformer: Prediction Meets Generation](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/regression-transformer/)
- [Exposing Limitations of Molecular ML with Activity Cliffs](https://hunterheidenreich.com/notes/chemistry/molecular-design/property-prediction/activity-cliffs-benchmark/)

#### Molecular Representations

##### Molecular Notations
- [InChI: The International Chemical Identifier](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/inchi/)
- [Materials Representations for ML Review](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/materials-representations-ml-review/)
- [Atom-in-SMILES: Better Tokens for Chemical Models](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/atom-in-smiles-tokenization/)
- [DeepSMILES: Adapting SMILES Syntax for Machine Learning](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/deepsmiles-adaptation-for-ml/)
- [Group SELFIES: Fragment-Based Molecular Strings](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/group-selfies-fragment-molecular-representation/)
- [Randomized SMILES Improve Molecular Generative Models](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/randomized-smiles-generative-models/)
- [SMI+AIS: Hybridizing SMILES with Environment Tokens](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/smi-ais-hybrid-molecular-representation/)
- [SMILES vs SELFIES Tokenization for Chemical LMs](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/smiles-selfies-tokenization-chemical-lm/)
- [Smirk: Complete Tokenization for Molecular Models](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/smirk-tokenization-molecular-models/)
- [SPE: Data-Driven SMILES Substructure Tokenization](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/smiles-pair-encoding/)
- [t-SMILES: Tree-Based Fragment Molecular Encoding](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/t-smiles-fragment-molecular-representation/)
- [Review of Molecular Representation Learning Models](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/molecular-representation-learning-foundation-models-review/)
- [Invalid SMILES Benefit Chemical Language Models: A Study](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/invalid-smiles-help/)
- [SELFIES and the Future of Molecular String Representations](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/selfies-2022/)
- [InChI and Tautomerism: Toward Comprehensive Treatment](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/inchi-and-tautomers/)
- [InChI: The Worldwide Chemical Structure Identifier Standard](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/inchi-2013/)
- [Making InChI FAIR and Sustainable for Inorganic Chemistry](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/inchi-2025/)
- [Mixfile & MInChI: Machine-Readable Mixture Formats](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/mixfile-minchi/)
- [NInChI: Toward a Chemical Identifier for Nanomaterials](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/ninchi-alpha/)
- [Recent Advances in the SELFIES Library: 2023 Update](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/selfies-2023/)
- [RInChI: The Reaction International Chemical Identifier](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/rinchi/)
- [SELFIES: The Original Paper on Robust Molecular Strings](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/selfies-original-paper/)
- [SMILES Notation: The Original Paper by Weininger (1988)](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/smiles-original-paper/)
- [SELFIES: A Robust Molecular String Representation](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/selfies/)
- [SMILES: A Compact Notation for Chemical Structures](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/smiles/)
- [The Number of Isomeric Hydrocarbons of the Methane Series](https://hunterheidenreich.com/notes/chemistry/molecular-representations/notations/number-of-isomeric-hydrocarbons/)

##### Multimodal Molecular Models
- [BioT5: Cross-Modal Integration of Biology and Chemistry](https://hunterheidenreich.com/notes/chemistry/molecular-representations/multimodal/biot5-cross-modal-biology/)
- [MolFM: Trimodal Molecular Foundation Pre-training](https://hunterheidenreich.com/notes/chemistry/molecular-representations/multimodal/molfm-multimodal-molecular-foundation/)
- [MoMu: Bridging Molecular Graphs and Natural Language](https://hunterheidenreich.com/notes/chemistry/molecular-representations/multimodal/momu-molecular-multimodal-foundation/)
- [DMP: Dual-View Molecule Pre-training (SMILES+GNN)](https://hunterheidenreich.com/notes/chemistry/molecular-representations/multimodal/dual-view-molecule-pretraining/)
- [MG-BERT: Graph BERT for Molecular Property Prediction](https://hunterheidenreich.com/notes/chemistry/molecular-representations/multimodal/mg-bert-molecular-graph-bert/)
- [nach0: A Multimodal Chemical and NLP Foundation Model](https://hunterheidenreich.com/notes/chemistry/molecular-representations/multimodal/nach0-multimodal-chemical-language-model/)
- [SPMM: A Bidirectional Molecular Foundation Model](https://hunterheidenreich.com/notes/chemistry/molecular-representations/multimodal/spmm-bidirectional-structure-property/)

##### Molecular Encoders
- [Mol2vec: Unsupervised ML with Chemical Intuition](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/mol2vec-unsupervised-chemical-intuition/)
- [CDDD: Learning Descriptors by Translating SMILES](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/cddd-translation-molecular-descriptors/)
- [MolBERT: Auxiliary Tasks for Molecular BERT Models](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/molbert-molecular-representations/)
- [Seq2seq Fingerprint: Unsupervised Molecular Embedding](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/seq2seq-fingerprint-molecular-embedding/)
- [SMI-TED: Encoder-Decoder Foundation Models for Chemistry](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/smi-ted-encoder-decoder-chemistry/)
- [SMILES Transformer: Low-Data Molecular Fingerprints](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/smiles-transformer/)
- [SMILES-BERT: BERT-Style Pre-Training for Molecules](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/smiles-bert/)
- [Survey of Transformer Architectures in Molecular Science](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/transformers-molecular-science-review/)
- [Systematic Review of Deep Learning CLMs (2020-2024)](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/systematic-review-deep-learning-clms/)
- [Transformer CLMs for SMILES: Literature Review 2024](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/transformer-clms-smiles-review/)
- [X-MOL: Pre-training on 1.1B Molecules for SMILES](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/x-mol-pretraining-molecular-understanding/)
- [AMORE: Testing ChemLLM Robustness to SMILES Variants](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/amore-smiles-robustness-framework/)
- [Neural Scaling of Deep Chemical Models](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/neural-scaling-of-deep-chemical-models/)
- [BARTSmiles: BART Pre-Training for Molecular SMILES](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/bartsmiles-molecular-representations/)
- [MoLFormer: Large-Scale Chemical Language Representations](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/molformer/)
- [SELFormer: A SELFIES-Based Molecular Language Model](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/selformer/)
- [ChemBERTa-3: Open Source Chemical Foundation Models](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/chemberta-3/)
- [ChemBERTa-2: Scaling Molecular Transformers to 77M](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/chemberta-2/)
- [ChemBERTa: Molecular Property Prediction via Transformers](https://hunterheidenreich.com/notes/chemistry/molecular-representations/encoders/chemberta/)

##### Chemical Name Translation
- [Neural Machine Translation of Chemical Nomenclature](https://hunterheidenreich.com/notes/chemistry/molecular-representations/name-translation/nmt-chemical-nomenclature-en-zh/)
- [Transformer Name-to-SMILES with Atom Count Losses](https://hunterheidenreich.com/notes/chemistry/molecular-representations/name-translation/transformer-chemical-name-to-smiles/)
- [STOUT V2.0: Transformer-Based SMILES to IUPAC Translation](https://hunterheidenreich.com/notes/chemistry/molecular-representations/name-translation/stout-v2/)
- [STOUT: SMILES to IUPAC Names via Neural Machine Translation](https://hunterheidenreich.com/notes/chemistry/molecular-representations/name-translation/stout/)
- [Struct2IUPAC: Translating SMILES to IUPAC via Transformers](https://hunterheidenreich.com/notes/chemistry/molecular-representations/name-translation/struct2iupac-2021/)
- [Translating InChI to IUPAC Names with Transformers](https://hunterheidenreich.com/notes/chemistry/molecular-representations/name-translation/handsel-inchi-iupac-2021/)

#### Optical Chemical Structure Recognition

##### Markush Structures
- [MarkushGrapher-2: End-to-End Markush Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/markush/markushgrapher-2-multimodal-recognition/)
- [MarkushGrapher: Multi-modal Markush Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/markush/markushgrapher/)
- [One Strike, You're Out: Detecting Markush Structures](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/markush/jurriaans-markush-detection-2023/)

##### Benchmarks and Reviews
- [Uni-Parser: Industrial-Grade Multi-Modal PDF Parsing (2025)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/uni-parser-2025/)
- [Image-to-Sequence OCSR: A Comparative Analysis](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/image-to-sequence-comparison/)
- [Benchmarking Eight OCSR Tools on Patent Images (2024)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/krasnov-ocsr-benchmark-2024/)
- [Review of OCSR Techniques and Models (Musazade 2022)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/musazade-ocsr-review-2022/)
- [String Representations for Chemical Image Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/rajan-string-representations-2022/)
- [A Review of Optical Chemical Structure Recognition Tools](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/rajan-ocsr-review-2020/)
- [Chemical Structure Reconstruction with chemoCR (2011)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/chemocr-trec-2011/)
- [ChemReader Image-to-Structure OCR at TREC 2011 Chemical IR](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/chemreader-trec-2011/)
- [CLEF-IP 2012: Patent and Chemical Structure Benchmark](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/clef-ip-2012/)
- [MolRec at CLEF 2012: Rule-Based Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/molrec-clef-2012/)
- [OSRA at CLEF-IP 2012: Native TIFF Processing for Patents](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/osra-clef-2012/)
- [Overview of the TREC 2011 Chemical IR Track Benchmark](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/trec-chem-2011/)
- [Imago: Open-Source Chemical Structure Recognition (2011)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/imago-trec-2011/)
- [OSRA at TREC-CHEM 2011: Optical Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/osra-trec-2011/)
- [OCSR Methods: A Taxonomy of Approaches](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/ocsr-methods/)
- [MolRec: Chemical Structure Recognition at CLEF 2012](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/molrec_at_clef/)
- [MolRec: Rule-Based OCSR System at TREC 2011 Benchmark](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/benchmarks/molrec_at_trec/)

##### Image-to-Graph Models
- [AdaptMol: Domain Adaptation for Molecular OCSR (2026)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/adaptmol-2026/)
- [GraSP: Graph Recognition via Subgraph Prediction (2026)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/grasp-2026/)
- [MolGrapher: Graph-based Chemical Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/molgrapher/)
- [MolMole: Unified Vision Pipeline for Molecule Mining](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/molmole/)
- [MolScribe: Robust Image-to-Graph Molecular Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/molscribe/)
- [ABC-Net: Keypoint-Based Molecular Image Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/abc-net/)
- [Image-to-Graph Transformers for Chemical Structures](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/image-to-graph-transformers/)
- [MolMiner: Deep Learning OCSR with YOLOv5 Detection](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/molminer/)
- [ChemGrapher: Deep Learning for Chemical Graph OCSR](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-graph/chemgrapher-2020/)

##### Rule-Based Systems
- [GraphReco: Probabilistic Structure Recognition (2026)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/graphreco-2026/)
- [Probabilistic OCSR with Markov Logic Networks](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/mlocsr/)
- [Research on Chemical Expression Images Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/hong-chemical-expression-2015/)
- [Chemical Structure Recognition (Rule-Based)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/molrec-2012/)
- [CLiDE Pro: Optical Chemical Structure Recognition Tool](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/clide-pro-2009/)
- [Kekulé-1 System for Chemical Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/kekule-1996/)
- [Automatic Recognition of Chemical Images](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/algorri-chemical-image-recognition-2007/)
- [Chemical Literature Data Extraction: The CLiDE Project](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/clide-1993/)
- [Chemical Machine Vision](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/chemical-machine-vision/)
- [ChemReader: Automated Structure Extraction](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/chemreader-2009/)
- [Graph Perception for Chemical Structure OCR](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/contreras-ocr-1990/)
- [Kekulé: OCR-Optical Chemical Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/kekule-1992/)
- [Optical Recognition of Chemical Graphics](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/casey-ocsr-1993/)
- [OSRA: Open Source Optical Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/osra/)
- [Reconstruction of Chemical Molecules from Images](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/algorri-reconstruction-2007/)
- [ChemInfty: Chemical Structure Recognition in Patent Images](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/rule-based/cheminfty/)

##### Vision-Language Models
- [GTR-CoT: Graph Traversal Chain-of-Thought for Molecules](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/vision-language/gtr-mol-vlm/)
- [OCSU: Optical Chemical Structure Understanding (2025)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/vision-language/ocsu/)
- [MolNexTR: A Dual-Stream Molecular Image Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/vision-language/molnextr/)
- [MolParser-7M & WildMol: Large-Scale OCSR Datasets](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/vision-language/molparser_7m-wildmol/)
- [MolParser: End-to-End Molecular Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/vision-language/mol-parser/)
- [SubGrapher: Visual Fingerprinting of Chemical Structures](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/vision-language/subgrapher/)

##### Hand-Drawn Structure Recognition
- [OCSAug: Diffusion-Based Augmentation for Hand-Drawn OCSR](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/ocsaug/)
- [AtomLenz: Atom-Level OCSR with Limited Supervision](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/atomlenz/)
- [ChemReco: Hand-Drawn Chemical Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/chemreco/)
- [Enhanced DECIMER for Hand-Drawn Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/decimer-hand-drawn/)
- [ChemPix: Hand-Drawn Hydrocarbon Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/chempix/)
- [Handwritten Chemical Structure Recognition with RCGD](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/hu-handwritten-rcgd-2023/)
- [Handwritten Chemical Ring Recognition with Neural Networks](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/hewahi-ring-recognition-2008/)
- [Structural Analysis of Handwritten Chemical Formulas](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/ramel-handwritten-1999/)
- [Hand-Drawn Chemical Diagram Recognition (AAAI 2007)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/hand-drawn/ouyang-davis-aaai-2007/)

##### Image-to-Sequence Models
- [DECIMER.ai: Optical Chemical Structure Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/decimer-ai/)
- [Dual-Path Global Awareness Transformer (DGAT) for OCSR](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/dgat/)
- [Image2InChI: SwinTransformer for Molecular Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/image2inchi/)
- [MMSSC-Net: Multi-Stage Sequence Cognitive Networks](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/mmssc-net/)
- [MolSight: OCSR with RL and Multi-Granularity Learning](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/molsight/)
- [DECIMER 1.0: Transformers for Chemical Image Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/decimer-1.0/)
- [End-to-End Transformer for Molecular Image Captioning](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/vit-inchi-transformer/)
- [ICMDT: Automated Chemical Structure Image Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/icmdt/)
- [Image2SMILES: Transformer OCSR with Synthetic Data Pipeline](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/image2smiles/)
- [MICER: Molecular Image Captioning with Transfer Learning](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/micer/)
- [SwinOCSR: End-to-End Chemical OCR with Swin Transformers](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/swinocsr/)
- [DECIMER: Deep Learning for Chemical Image Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/decimer/)
- [Deep Learning for Molecular Structure Extraction (2019)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/staker-deep-learning-2019/)
- [Img2Mol: Accurate SMILES Recognition from Depictions](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/img2mol/)
- [IMG2SMI: Translating Molecular Structure Images to SMILES](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/img2smi/)
- [αExtractor: Chemical Info from Biomedical Literature](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/alpha-extractor/)
- [RFL: Simplifying Chemical Structure Recognition (AAAI 2025)](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/image-to-sequence/rfl/)

##### Online Recognition
- [Handwritten Chemical Symbol Recognition Using SVMs](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/tang-online-symbol-2013/)
- [HMM-based Online Recognition of Chemical Symbols](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/zhang-hmm-handwriting-2009/)
- [On-line Handwritten Chemical Expression Recognition](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/yang-icpr-2008/)
- [Online Handwritten Chemical Formula Structure Analysis](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/wang-online-handwritten-2009/)
- [Recognition of On-line Handwritten Chemical Expressions](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/yang-online-handwritten-2008/)
- [SVM-HMM Online Classifier for Chemical Symbols](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/zhang-svm-hmm-2010/)
- [Unified Framework for Handwritten Chemical Expressions](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/chang-unified-framework-2009/)
- [ChemInk: Real-Time Recognition for Chemical Drawings](https://hunterheidenreich.com/notes/chemistry/optical-structure-recognition/online-recognition/chemink-2011/)

#### LLMs for Chemistry
- [ChatDrug: Conversational Drug Editing with ChatGPT](https://hunterheidenreich.com/notes/chemistry/llm-applications/chatdrug-conversational-drug-editing/)
- [ChemCrow: Augmenting LLMs with 18 Chemistry Tools](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemcrow-augmenting-llms-chemistry-tools/)
- [ChemLLM: A Chemical Large Language Model Framework](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemllm-chemical-large-language-model/)
- [Coscientist: Autonomous Chemistry with LLM Agents](https://hunterheidenreich.com/notes/chemistry/llm-applications/autonomous-chemical-research-coscientist/)
- [DrugAssist: Interactive LLM Molecule Optimization](https://hunterheidenreich.com/notes/chemistry/llm-applications/drugassist-llm-molecule-optimization/)
- [DrugChat: Conversational QA on Drug Molecule Graphs](https://hunterheidenreich.com/notes/chemistry/llm-applications/drugchat-chatgpt-drug-molecule-graphs/)
- [Fine-Tuning GPT-3 for Predictive Chemistry Tasks](https://hunterheidenreich.com/notes/chemistry/llm-applications/leveraging-llms-predictive-chemistry/)
- [Galactica: A Curated Scientific LLM from Meta AI](https://hunterheidenreich.com/notes/chemistry/llm-applications/galactica-large-language-model-for-science/)
- [LlaSMol: Instruction-Tuned LLMs for Chemistry Tasks](https://hunterheidenreich.com/notes/chemistry/llm-applications/llamsmol-instruction-tuning-chemistry/)
- [PharmaGPT: Domain-Specific LLMs for Pharma and Chem](https://hunterheidenreich.com/notes/chemistry/llm-applications/pharmagpt-domain-specific-llms-biopharmaceutical/)
- [Transformers and LLMs for Chemistry Drug Discovery](https://hunterheidenreich.com/notes/chemistry/llm-applications/transformers-llms-chemistry-drug-discovery/)
- [ChemLLMBench: Benchmarking LLMs on Chemistry Tasks](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemllmbench-eight-chemistry-tasks/)
- [Fine-Tuning GPT-3 for Molecular Property Prediction](https://hunterheidenreich.com/notes/chemistry/llm-applications/fine-tuning-gpt3-molecular-properties/)
- [Foundation Models in Chemistry: A 2025 Perspective](https://hunterheidenreich.com/notes/chemistry/llm-applications/foundation-models-chemistry-perspective/)
- [LLM4Mol: ChatGPT Captions as Molecular Representations](https://hunterheidenreich.com/notes/chemistry/llm-applications/llm4mol-captions-as-representations/)
- [MaCBench: Multimodal Chemistry and Materials Benchmark](https://hunterheidenreich.com/notes/chemistry/llm-applications/macbench-multimodal-chemistry-benchmark/)
- [NLP Models That Automate Programming for Chemistry](https://hunterheidenreich.com/notes/chemistry/llm-applications/nlp-models-transform-chemistry/)
- [Survey of Scientific LLMs in Bio and Chem Domains](https://hunterheidenreich.com/notes/chemistry/llm-applications/scientific-llm-survey-bio-chem/)
- [Benchmarking Chemistry Knowledge in Code-Gen LLMs](https://hunterheidenreich.com/notes/chemistry/llm-applications/llm-chemistry-code-assessment/)
- [Benchmarking LLMs for Molecular Property Prediction](https://hunterheidenreich.com/notes/chemistry/llm-applications/benchmarking-llms-molecule-prediction/)
- [ChemBench: Evaluating LLM Chemistry Against Experts](https://hunterheidenreich.com/notes/chemistry/llm-applications/chembench-llm-chemistry-evaluation/)
- [ChemEval: Fine-Grained LLM Evaluation for Chemistry](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemeval-multilevel-chemical-evaluation/)
- [ChemSafetyBench: Benchmarking LLM Safety in Chemistry](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemsafetybench-llm-safety-chemistry/)
- [ChemDFM-R: Chemical Reasoning LLM with Atomized Knowledge](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemdfm-r/)
- [ChemDFM-X: Multimodal Foundation Model for Chemistry](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemdfm-x/)
- [InstructMol: Multi-Modal Molecular LLM for Drug Discovery](https://hunterheidenreich.com/notes/chemistry/llm-applications/instructmol/)
- [MERMaid: Multimodal Chemical Reaction Mining from PDFs](https://hunterheidenreich.com/notes/chemistry/llm-applications/mermaid/)
- [Multimodal Search in Chemical Documents and Reactions](https://hunterheidenreich.com/notes/chemistry/llm-applications/shah-multimodal-search-2025/)
- [ChemVLM: A Multimodal Large Language Model for Chemistry](https://hunterheidenreich.com/notes/chemistry/llm-applications/chemvlm/)

### Machine Learning

#### Geometric Deep Learning
- [Defining Disentangled Representations via Group Theory](https://hunterheidenreich.com/notes/machine-learning/geometric-deep-learning/defining-disentangled-representations/)
- [DGCNN: Dynamic Graph CNN for Point Cloud Learning](https://hunterheidenreich.com/notes/machine-learning/geometric-deep-learning/dgcnn-dynamic-graph-point-clouds/)
- [SE(3)-Transformers: Equivariant Attention for 3D Data](https://hunterheidenreich.com/notes/machine-learning/geometric-deep-learning/se3-transformers/)
- [Spherical CNNs: Rotation-Equivariant Networks on the Sphere](https://hunterheidenreich.com/notes/machine-learning/geometric-deep-learning/spherical-cnns/)
- [3D Steerable CNNs: Rotationally Equivariant Features](https://hunterheidenreich.com/notes/machine-learning/geometric-deep-learning/3d-steerable-cnns/)

#### Model Architectures
- [Lagrangian Neural Networks for Physics](https://hunterheidenreich.com/notes/machine-learning/model-architectures/lagrangian-neural-networks/)
- [Liquid-S4: Input-Dependent State-Space Models](https://hunterheidenreich.com/notes/machine-learning/model-architectures/liquid-s4-state-space-models/)
- [NaViT: Native Resolution Vision Transformer](https://hunterheidenreich.com/notes/machine-learning/model-architectures/navit-native-resolution-vit/)
- [Can Recurrent Neural Networks Warp Time? (ICLR 2018)](https://hunterheidenreich.com/notes/machine-learning/model-architectures/can-recurrent-neural-networks-warp-time/)
- [Relational Inductive Biases in Deep Learning (2018)](https://hunterheidenreich.com/notes/machine-learning/model-architectures/relational-inductive-biases-deep-learning-graph-networks/)
- [Scaling Laws vs Model Architectures: Inductive Bias](https://hunterheidenreich.com/notes/machine-learning/model-architectures/scaling-laws-vs-model-architectures/)
- [The Quarks of Attention: Building Blocks of Attention](https://hunterheidenreich.com/notes/machine-learning/model-architectures/quarks-of-attention/)
- [Distributed Representations: A Foundational Theory](https://hunterheidenreich.com/notes/machine-learning/model-architectures/distributed-representations/)
- [Communication in the Presence of Noise: Shannon's 1949 Paper](https://hunterheidenreich.com/notes/machine-learning/model-architectures/communication-in-the-presence-of-noise/)

#### Generative Models
- [Consistency Models: Fast One-Step Diffusion Generation](https://hunterheidenreich.com/notes/machine-learning/generative-models/consistency-models/)
- [D3PM: Discrete Denoising Diffusion Probabilistic Models](https://hunterheidenreich.com/notes/machine-learning/generative-models/discrete-diffusion-models/)
- [Latent Diffusion Models for High-Res Image Synthesis](https://hunterheidenreich.com/notes/machine-learning/generative-models/latent-diffusion-models/)
- [A Convexity Principle for Interacting Gases (McCann 1997)](https://hunterheidenreich.com/notes/machine-learning/generative-models/convexity-principle-interacting-gases/)
- [Building Normalizing Flows with Stochastic Interpolants](https://hunterheidenreich.com/notes/machine-learning/generative-models/stochastic-interpolants/)
- [Flow Matching for Generative Modeling: Scalable CNFs](https://hunterheidenreich.com/notes/machine-learning/generative-models/flow-matching-for-generative-modeling/)
- [Neural ODEs: Continuous-Depth Deep Learning Models](https://hunterheidenreich.com/notes/machine-learning/generative-models/neural-odes/)
- [Rectified Flow: Learning to Generate and Transfer Data](https://hunterheidenreich.com/notes/machine-learning/generative-models/rectified-flow/)
- [Score Matching and Denoising Autoencoders: A Connection](https://hunterheidenreich.com/notes/machine-learning/generative-models/score-matching-denoising-autoencoders/)
- [Score-Based Generative Modeling with SDEs (Song 2021)](https://hunterheidenreich.com/notes/machine-learning/generative-models/score-based-generative-modeling-sde/)
- [Mixture Density Networks: Modeling Multimodal Distributions](https://hunterheidenreich.com/notes/machine-learning/generative-models/mixture-density-networks/)
- [Auto-Encoding Variational Bayes: VAE Paper Summary](https://hunterheidenreich.com/notes/machine-learning/generative-models/autoencoding-variational-bayes/)
- [Importance Weighted Autoencoders (IWAE) for Tighter Bounds](https://hunterheidenreich.com/notes/machine-learning/generative-models/importance-weighted-autoencoders/)
- [Contrastive Learning for Variational Autoencoder Priors](https://hunterheidenreich.com/notes/machine-learning/generative-models/contrastive-learning-for-vae-priors/)

### Natural Language Processing

#### Language Models
- [SpeechT5: Unified Speech-Text Pre-Training Framework](https://hunterheidenreich.com/notes/natural-language-processing/language-models/speecht5-unified-speech-text-pretraining/)
- [Data Mixing Laws for LM Pretraining Optimization](https://hunterheidenreich.com/notes/natural-language-processing/language-models/data-mixing-laws-pretraining/)
- [DoReMi: Optimizing Data Mixtures for LM Pretraining](https://hunterheidenreich.com/notes/natural-language-processing/language-models/doremi-data-mixture-optimization/)
- [Scaling Data-Constrained Language Models](https://hunterheidenreich.com/notes/natural-language-processing/language-models/scaling-data-constrained-language-models/)
- [SlimPajama-DC: Data Combinations for LLM Training](https://hunterheidenreich.com/notes/natural-language-processing/language-models/slimpajama-dc-data-combinations/)
- [T5: Exploring Transfer Learning Limits](https://hunterheidenreich.com/notes/natural-language-processing/language-models/t5-text-to-text-transfer-transformer/)
- [Block-Recurrent Transformers for Long Sequences](https://hunterheidenreich.com/notes/natural-language-processing/language-models/block-recurrent-transformers/)
- [RWKV: Linear-Cost RNN with Transformer Training](https://hunterheidenreich.com/notes/natural-language-processing/language-models/rwkv-rnn-transformer-architecture/)

### Computational Biology
- [Arun et al.: SVD-Based Least-Squares Fitting of 3D Points](https://hunterheidenreich.com/notes/computational-biology/arun-svd-point-fitting/)
- [Horn et al.: Absolute Orientation Using Orthonormal Matrices](https://hunterheidenreich.com/notes/computational-biology/horn-orthonormal-matrices/)
- [Umeyama's Method: Corrected SVD for Point Alignment](https://hunterheidenreich.com/notes/computational-biology/umeyama-similarity-transformation/)
- [Horn's Method: Absolute Orientation via Unit Quaternions](https://hunterheidenreich.com/notes/computational-biology/horn-absolute-orientation/)
- [Kabsch Algorithm: Optimal Rotation for Point Set Alignment](https://hunterheidenreich.com/notes/computational-biology/kabsch-algorithm/)
- [DynamicFlow: Integrating Protein Dynamics into Drug Design](https://hunterheidenreich.com/notes/computational-biology/dynamicflow/)
- [How to Fold Graciously: Levinthal's Paradox (1969)](https://hunterheidenreich.com/notes/computational-biology/fold-graciously/)

### Scientific Research Methods & Frameworks
- [AI & Physical Sciences Taxonomy: A Seven-Vector Framework](https://hunterheidenreich.com/notes/research-methods/ai-physical-sciences-paper-taxonomy/)

## Videos
- [Müller-Brown Basin MA: Langevin Dynamics Simulation](https://hunterheidenreich.com/videos/muller-brown-basin-ma-simulation/)
- [Müller-Brown Basin MB: Langevin Dynamics Simulation](https://hunterheidenreich.com/videos/muller-brown-basin-mb-simulation/)
- [Müller-Brown Transition: Langevin Dynamics Simulation](https://hunterheidenreich.com/videos/muller-brown-transition-simulation/)
- [Liquid Argon: LAMMPS Simulation](https://hunterheidenreich.com/videos/liquid-argon-lammps-simulation/)
- [Copper Adatom Diffusion on Cu(100): LAMMPS Simulation](https://hunterheidenreich.com/videos/cu-adatom-diffusion/)
- [Platinum Adatom Diffusion on Pt(100): LAMMPS Simulation](https://hunterheidenreich.com/videos/pt-adatom-diffusion/)
- [Rubik's Cube Player - Drexel Music Hackathon 2017](https://hunterheidenreich.com/videos/rubiks-cube-player-hackathon/)
- [Internal Reflection - Audio Reactive Music Video](https://hunterheidenreich.com/videos/internal-reflection-music-video/)
- [Carnal Desires (My Soul) - Music Video](https://hunterheidenreich.com/videos/carnal-desires-music-video/)
- [Concavity - Music Video](https://hunterheidenreich.com/videos/concavity-music-video/)
- [False Gods - Official Music Video](https://hunterheidenreich.com/videos/false-gods-official-video/)
- [False Idols - Music Video](https://hunterheidenreich.com/videos/false-idols-music-video/)
- [In the Grass By a Lake - Music Video](https://hunterheidenreich.com/videos/in-the-grass-by-a-lake-music-video/)
- [The Secrecy in Dreaming - Music Video](https://hunterheidenreich.com/videos/the-secrecy-in-dreaming-music-video/)

## Leaderboards
- [US Policy Area Classification Leaderboard](https://hunterheidenreich.com/leaderboards/policy_area_classification_leaderboard/)