Smiles
PulseAugur coverage of Smiles — every cluster mentioning Smiles across labs, papers, and developer communities, ranked by signal.
7 day(s) with sentiment data
-
New molecular LLM uses substructure rationale for improved property prediction
Researchers have developed MR-MoL, a novel molecular large language model (LLM) designed for property prediction in drug discovery. Unlike existing models that represent molecules implicitly, MR-MoL explicitly incorpora…
-
MolBioKG system grounds unregistered molecules in biomedical knowledge graphs
Researchers have developed MolBioKG, a novel two-layer system designed to connect unregistered molecules to biomedical knowledge graphs. This system addresses the challenge of 'out-of-graph molecules' by using multi-res…
-
MinerU.Chem system achieves 93% accuracy in chemical structure recognition
A new system called MinerU.Chem has been developed to extract chemical structures and reactions from organic chemistry documents, converting them into machine-readable data. This system, integrated into the MinerU platf…
-
New model CheMatE unifies chemical structures and natural language
Researchers have developed CheMatE, a new embedding model designed to jointly represent chemical structures (SMILES) and natural language within a unified space. Built on a ModernBERT backbone, CheMatE employs a two-sta…
-
New SIGMA objective improves molecular autoregressive models
Researchers have developed SIGMA, a novel objective for autoregressive molecular models that improves their ability to assign probabilities to molecules regardless of their serialization format. This method uses a dense…
-
New generative model InVirtuoGen advances fragment-based drug discovery
Researchers have developed InVirtuoGen, a novel discrete flow generative model designed for fragment-based drug discovery. This model shifts the generation paradigm from completion to refinement, allowing for more effec…
-
New Python library bridges molecular ML and scikit-learn
A new Python library called scikit-fingerprints has been released, designed to integrate molecular machine learning functionalities with the scikit-learn ecosystem. This library, built upon RDKit, provides a unified int…
-
New Q-Steer method boosts molecular optimization for AI language models
Researchers have introduced Q-Steer, a novel method designed to enhance molecular policy optimization for language models. This technique addresses the challenge of delayed feedback in molecular generation by estimating…
-
OLEDLM: New Language Model for OLED Molecular Design
Researchers have developed OLEDLM, a novel language model specifically designed for the design of organic light-emitting diode (OLED) molecules. This model utilizes a LLaMA-style transformer architecture as a foundation…
-
S1-Omni model unifies scientific AI tasks, outperforms GPT-5.5 and Gemini-3.1 Pro
Researchers have introduced S1-Omni, a novel unified multimodal reasoning model designed to advance AI for Science (AI4S). This model addresses the fragmentation in current AI4S approaches by integrating the joint model…
-
New framework MolBasic enhances LLMs' molecular understanding via SMILES-Graph translation
Researchers have introduced MolBasic, a new framework designed to enhance the molecular understanding capabilities of large language models (LLMs). This approach addresses the issue of LLMs failing to reliably capture m…
-
New neural network fuses 3D geometry, topology, and physics for molecular prediction
Researchers have developed a novel Tri-Branch Modular Fusion Neural Network designed to improve molecular property prediction. This framework integrates three distinct data modalities: 3D spatial geometry using SchNet, …
-
BPE vs. Unigram-LM: Tokenization algorithms create distinct vocabularies for chemistry SMILES
A new research paper explores the differences between two common tokenization methods, byte-pair encoding (BPE) and Unigram-LM, when applied to chemical SMILES strings. The study found that these algorithms produce sign…
-
Rust library enhances LLM molecular reasoning with explicit graph formats
A Rust cheminformatics library called chematic has been developed to improve how Large Language Models (LLMs) process molecular data. The library addresses limitations of using simple SMILES strings by incorporating exp…
-
New FisherSketch method analyzes LLM update geometry at scale
Researchers have developed FisherSketch, a novel method for analyzing the geometry of updates in large language models (LLMs) with shared vocabularies. This technique allows for training-free source selection in scienti…
-
Chemical language models' internal representations analyzed with sparse autoencoders
A new research paper explores the internal workings of chemical language models (cLMs) by applying sparse autoencoders (SAEs) to MolFormer. The study reveals that early layers of the model focus on syntactic patterns an…
-
BioMatrix integrates sequences, structures, and language in new multimodal foundation model
Researchers have developed BioMatrix, a novel multimodal foundation model designed to integrate biological data types like sequences, structures, and natural language within a single architecture. Unlike previous models…
-
New benchmark MolGraphBench evaluates GNNs for molecular regression tasks
A new benchmark called MolGraphBench has been introduced to evaluate Graph Neural Network (GNN) architectures for molecular regression tasks. The benchmark, proposed by Ishaan Gupta, analyzes four common GNN models, fin…
-
LLM molecular tasks depend on representation, study finds
A new study on arXiv benchmarks the performance of 16 large language models across nine molecular representations for eight chemical tasks. The research found that model performance is heavily dependent on the molecular…
-
New GFlowNet training method improves LLM prefix balance and diversity
Researchers have introduced a new training method for Generative Flow Networks (GFlowNets) called Rooted absorbed prefix Trajectory Balance (RapTB), designed to address issues like prefix collapse and length bias in lar…