Selected work

Representative Papers

Uncertainty-aware reinforcement learning for chemical language models

Jun 23, 2026

This work addresses a critical limitation in conventional reinforcement learning for molecular generation: the neglect of uncertainty in property prediction, which often leads to unstable optimization and the proposal of spuriously high-scoring molecules. To remedy this, the study introduces the first reinforcement learning framework that explicitly incorporates predictive uncertainty into a chemical language model via a dual-path integration mechanism. Specifically, uncertainty is treated both as an auxiliary optimization objective to balance performance and reliability, and as a modulation signal for policy updates to downweight unreliable samples. By integrating ChemProp, random forests, and conformal prediction techniques, the proposed method maintains competitive molecular scores while substantially improving empirical validity—raising the true hit rate from 0.5 to 0.75 and nearly doubling the number of effective molecules generated.

0 citationsRead paper

AI-Guided Discovery of Novel Ionic Liquid Solvents for Industrial CO2 Capture

Jan 02, 2026

This study addresses CO₂ capture from industrial flue gas by proposing an AI-driven, end-to-end design framework for ionic liquid solvents as sustainable alternatives to energy-intensive and corrosive amine-based systems. The approach generates candidate molecules through combinatorial pairing of cations and anions, employs graph neural networks to predict CO₂ solubility and viscosity, and utilizes the Van’t Hoff model to estimate working capacity and regeneration energy. Integrated Pareto-based multi-objective optimization and synthetic feasibility analysis enable a closed-loop pipeline encompassing molecular generation, property prediction, performance optimization, and synthesizability validation. The method successfully identifies 36 high-performance ionic liquid candidates, projected to reduce operational costs by 5–10% and capital expenditures by up to 10%.

0 citationsRead paper

MSDM: Generating Task-Specific Pathology Images with a Multimodal Conditioned Diffusion Model for Cell and Nuclei Segmentation

Oct 10, 2025

Cell/nucleus segmentation in histopathological images suffers from severe annotation scarcity—particularly for rare morphologies—hindering robust model training. Method: We propose a novel multimodal conditional diffusion model that jointly integrates morphological maps (horizontal/vertical), RGB color features, and BERT-encoded textual metadata via multi-head cross-attention, enabling fine-grained, controllable image-mask pair generation. Contribution/Results: The method supports task-oriented data augmentation by synthesizing pixel-accurate, high-fidelity masks whose embedded feature distributions closely match those of real data (low Wasserstein distance). Experiments demonstrate substantial improvements in segmentation generalization—especially for underrepresented cell types such as columnar cells—validating the efficacy and novelty of multimodal conditional generation for computational pathology data augmentation.

0 citationsRead paper

Accelerated Medicines Development using a Digital Formulator and a Self-Driving Tableting DataFactory

Mar 20, 2025

Traditional tablet development is time-consuming and resource-intensive, impeding efficient prediction and optimization of Critical Quality Attributes (CQAs). To address this, we propose an intelligent pharmaceutical platform integrating digital prescription design with a self-driving tablet manufacturing data factory, enabling a fully automated, closed-loop workflow—from raw material characterization to qualified tablet production. The platform introduces a hybrid mechanistic–data-driven modeling approach for the digital prescriber and incorporates Bayesian optimization alongside fully automated powder feeding, compaction, and real-time in-line performance testing. It reduces formulation development for a single tablet to under six hours using less than 5 g of active pharmaceutical ingredient (API), and completes small-batch production of up to 1,440 tablets within 24 hours. Validation across multiple APIs and drug loadings achieves 100% success rate. This framework significantly enhances development efficiency and resource utilization, establishing a scalable paradigm for accelerated formulation development.

0 citationsRead paper

Recent publications

Latest Papers

Uncertainty-aware reinforcement learning for chemical language models

Jun 23, 2026

This work addresses a critical limitation in conventional reinforcement learning for molecular generation: the neglect of uncertainty in property prediction, which often leads to unstable optimization and the proposal of spuriously high-scoring molecules. To remedy this, the study introduces the first reinforcement learning framework that explicitly incorporates predictive uncertainty into a chemical language model via a dual-path integration mechanism. Specifically, uncertainty is treated both as an auxiliary optimization objective to balance performance and reliability, and as a modulation signal for policy updates to downweight unreliable samples. By integrating ChemProp, random forests, and conformal prediction techniques, the proposed method maintains competitive molecular scores while substantially improving empirical validity—raising the true hit rate from 0.5 to 0.75 and nearly doubling the number of effective molecules generated.

0 citationsRead paper

AI-Guided Discovery of Novel Ionic Liquid Solvents for Industrial CO2 Capture

Jan 02, 2026

This study addresses CO₂ capture from industrial flue gas by proposing an AI-driven, end-to-end design framework for ionic liquid solvents as sustainable alternatives to energy-intensive and corrosive amine-based systems. The approach generates candidate molecules through combinatorial pairing of cations and anions, employs graph neural networks to predict CO₂ solubility and viscosity, and utilizes the Van’t Hoff model to estimate working capacity and regeneration energy. Integrated Pareto-based multi-objective optimization and synthetic feasibility analysis enable a closed-loop pipeline encompassing molecular generation, property prediction, performance optimization, and synthesizability validation. The method successfully identifies 36 high-performance ionic liquid candidates, projected to reduce operational costs by 5–10% and capital expenditures by up to 10%.

0 citationsRead paper

MSDM: Generating Task-Specific Pathology Images with a Multimodal Conditioned Diffusion Model for Cell and Nuclei Segmentation

Oct 10, 2025

Cell/nucleus segmentation in histopathological images suffers from severe annotation scarcity—particularly for rare morphologies—hindering robust model training. Method: We propose a novel multimodal conditional diffusion model that jointly integrates morphological maps (horizontal/vertical), RGB color features, and BERT-encoded textual metadata via multi-head cross-attention, enabling fine-grained, controllable image-mask pair generation. Contribution/Results: The method supports task-oriented data augmentation by synthesizing pixel-accurate, high-fidelity masks whose embedded feature distributions closely match those of real data (low Wasserstein distance). Experiments demonstrate substantial improvements in segmentation generalization—especially for underrepresented cell types such as columnar cells—validating the efficacy and novelty of multimodal conditional generation for computational pathology data augmentation.

0 citationsRead paper

Accelerated Medicines Development using a Digital Formulator and a Self-Driving Tableting DataFactory

Mar 20, 2025

Traditional tablet development is time-consuming and resource-intensive, impeding efficient prediction and optimization of Critical Quality Attributes (CQAs). To address this, we propose an intelligent pharmaceutical platform integrating digital prescription design with a self-driving tablet manufacturing data factory, enabling a fully automated, closed-loop workflow—from raw material characterization to qualified tablet production. The platform introduces a hybrid mechanistic–data-driven modeling approach for the digital prescriber and incorporates Bayesian optimization alongside fully automated powder feeding, compaction, and real-time in-line performance testing. It reduces formulation development for a single tablet to under six hours using less than 5 g of active pharmaceutical ingredient (API), and completes small-batch production of up to 1,440 tablets within 24 hours. Validation across multiple APIs and drug loadings achieves 100% success rate. This framework significantly enhances development efficiency and resource utilization, establishing a scalable paradigm for accelerated formulation development.

0 citationsRead paper