Deconstructing Neural Networks for Time Series Forecasting
Ablation study of neural network components for forecasting, finding gating and attention improve RNNs while recurrence …...
Transformers
GTR-CoT: Graph Traversal as Visual Chain of Thought for Molecular Structure Recognition
Wang et al.'s novel VLM approach using graph traversal CoT and faithful recognition principles to improve optical …...
αExtractor: Automatic Chemical Information Extraction from Biomedical Literature
Xiong et al.'s deep learning system for extracting chemical structures from literature images using ResNet-Transformer …...
MolNexTR: A Generalized Deep Learning Model for Molecular Image Recognition
Chen et al.'s dual-stream encoder approach for robust molecular structure recognition from diverse real-world images …...
MolParser: End-to-end Visual Recognition of Molecule Structures in the Wild
Fang et al.'s method for converting molecular structure images from scientific documents into machine-readable formats …...
Beyond Atoms: Enhancing Molecular Pretrained Representations with 3D Space Modeling
Lu et al. introduce SpaceFormer, a Transformer that models entire 3D molecular space—not just atoms—for superior …...
Invalid SMILES are Beneficial Rather than Detrimental to Chemical Language Models
Skinnider's 2024 Nature Machine Intelligence paper demonstrates that the ability to generate invalid SMILES is actually …...
Sarcasm Detection with Transformers: A Cautionary Tale
Learn how dataset bias can lead to misleading results in NLP: a sarcasm detection model that actually learned to …
IMG2SMI: Translating Molecular Structure Images to SMILES
Campos & Ji's method for converting 2D molecular images to SMILES strings using Transformers and SELFIES representation....
PyConversations: Social Media Conversational Analysis
Undergraduate thesis exploring representation learning for social media text and developing tools for cross-platform …
CoQA Dataset: Advancing Conversational Question Answering
Analysis of CoQA, a conversational QA dataset with multi-turn dialogue, coreference resolution, and natural answers for …