arXiv AI
CoMPASS: Collaborative Molecular Property Prediction via Adaptive Small-Large Model Synergy
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv Machine Learning
Aug 20
Monroe: A Molecular Foundation Model for In-Context Probabilistic Inference
Monroe is a new molecular foundation model that improves upon existing models by pre‑training on over 81 million molecules from the PM6 quantum chemistry dataset, enhancing stereochemistry representation, and introducing novel training losses such as conformer denoising and embedding decorrelation. It also incorporates a prior‑data‑fitted model (TabPFN) for downstream in‑context prediction and demonstrates superior performance on Polaris benchmarks and activity cliff tests. Ablation studies show that the PFN‑based downstream approach can upgrade other models, producing state‑of‑the‑art variants MiniMol_PFN and CheMeleon_PFN.
arXiv Machine Learning
Sep 17
A Multitask Large Reasoning Model for Molecular Science
The paper introduces a multitask large reasoning model for molecular science that incorporates chemical knowledge via a multispecialist architecture, chain-of-thought supervision, and molecule-informed reinforcement learning. It coordinates prediction and inference specialists across ten molecular tasks—including description, generation, nomenclature translation, property prediction, and reaction prediction—using task-conditioned routing. The model surpasses more than 20 general-purpose and molecular large language models, improving aggregate performance by 50.3% and outperforming leading multitask baselines on most tasks, while maintaining interpretable chemical inference and demonstrating a workflow for CNS candidate generation and retrosynthetic planning.
arXiv Machine Learning
Sep 7
Training Large Language Models for Small-Molecule Design with Synthetic Task Scaling
The paper explores how large language models (LLMs) can be trained for small-molecule drug design by using synthetic tasks that are cheaper to evaluate. By employing a curriculum that gradually increases task difficulty, the authors demonstrate that LLMs can learn design strategies that outperform larger models on structure-based lead optimization. This approach shows that scaling post‑training with synthetic tasks can effectively adapt LLMs to high‑cost experimental scenarios that are otherwise infeasible to train on directly.
arXiv Machine Learning
Aug 27
VINCENT: Validated Interaction Network for Cross-drug Explanation of Therapeutics
VINCENT is a post‑training framework that provides validated, chemically coherent explanations for drug synergy predictions by extracting atom‑pair evidence from attention and gradient signals, grouping them into motifs, and refining these motifs through repeated local perturbations. On a literature‑annotated subset of 25 drug pairs, VINCENT achieves a mean motif recall of 0.826, outperforming baselines (0.49–0.66). Across 71 test pairs, its validated interaction scores yield a TP/TN separation of 3.36, indicating more accurate recovery of literature‑supported molecular regions and better alignment with predictor behavior.
Hugging Face Trending Papers
Aug 26
VINCENT: Validated Interaction Network for Cross-drug Explanation of Therapeutics
Drug synergy prediction estimates whether two drugs produce a stronger joint effect than expected from their individual activities. For drug combination discovery, a single synergy score is often not...
arXiv Computation and Language
5d ago
Beyond Benchmark Accuracy: A Diagnostic Test for LLM Chemical Prediction
arXiv:2610.19906v1 Announce Type: new Abstract: LLMs have achieved strong performance on chemical prediction benchmarks. However, existing evaluation methods typically provide target entity identity...