Selected work

Representative Papers

CARE: Controlling LLM-Generated Policies through Auditable Review of Evidence in Scientific Experimentation

Jun 12, 2026

This work addresses the challenge of safely harnessing large language models (LLMs) in high-throughput experimental optimization, where direct LLM use risks unsafe exploration yet complete exclusion forfeits their optimization potential. To reconcile this trade-off, the authors propose the CARE framework, which employs a non-LLM default optimizer as the primary pathway while leveraging the LLM to generate candidate strategies. Adoption of these candidates is governed by an evidence-based intervention gating mechanism that audits proposals against publicly available evidence, ensuring decisions are auditable, controllable, and traceable. By synergistically integrating LLM-driven creativity with evidence-guided safety constraints, CARE achieves state-of-the-art performance on the Minerva/Olympus and ChemLex benchmarks, improving peak scores from 80.0 to 88.5 and from 83.9 to 92.1, respectively.

0 citationsRead paper

OCOO-T : A Simple and Scalable Virtual Cell Model for Transcriptional Perturbation Response Prediction

Jun 10, 2026

Predicting the transcriptional responses of single cells to genetic, chemical, and cytokine perturbations remains a central challenge in computational biology. This work proposes OCOO-T, a minimalist in silico cell model based on flow matching that directly operates on continuous gene expression profiles using a standard Transformer architecture. OCOO-T formulates perturbation response modeling as a continuous-time denoising process and integrates perturbation type, dosage, and cell-type information through adaptive layer normalization and contextual token fusion. By eschewing complex encoders or predefined structural priors, the method achieves markedly improved scalability and generalization. Evaluated on the Tahoe100M, Replogle, and PBMC benchmarks, OCOO-T demonstrates state-of-the-art performance, effectively generalizing across diverse perturbations and cell lines while efficiently handling high-dimensional transcriptomic data.

0 citationsRead paper

Recent publications

Latest Papers

CARE: Controlling LLM-Generated Policies through Auditable Review of Evidence in Scientific Experimentation

Jun 12, 2026

This work addresses the challenge of safely harnessing large language models (LLMs) in high-throughput experimental optimization, where direct LLM use risks unsafe exploration yet complete exclusion forfeits their optimization potential. To reconcile this trade-off, the authors propose the CARE framework, which employs a non-LLM default optimizer as the primary pathway while leveraging the LLM to generate candidate strategies. Adoption of these candidates is governed by an evidence-based intervention gating mechanism that audits proposals against publicly available evidence, ensuring decisions are auditable, controllable, and traceable. By synergistically integrating LLM-driven creativity with evidence-guided safety constraints, CARE achieves state-of-the-art performance on the Minerva/Olympus and ChemLex benchmarks, improving peak scores from 80.0 to 88.5 and from 83.9 to 92.1, respectively.

0 citationsRead paper

OCOO-T : A Simple and Scalable Virtual Cell Model for Transcriptional Perturbation Response Prediction

Jun 10, 2026

Predicting the transcriptional responses of single cells to genetic, chemical, and cytokine perturbations remains a central challenge in computational biology. This work proposes OCOO-T, a minimalist in silico cell model based on flow matching that directly operates on continuous gene expression profiles using a standard Transformer architecture. OCOO-T formulates perturbation response modeling as a continuous-time denoising process and integrates perturbation type, dosage, and cell-type information through adaptive layer normalization and contextual token fusion. By eschewing complex encoders or predefined structural priors, the method achieves markedly improved scalability and generalization. Evaluated on the Tahoe100M, Replogle, and PBMC benchmarks, OCOO-T demonstrates state-of-the-art performance, effectively generalizing across diverse perturbations and cell lines while efficiently handling high-dimensional transcriptomic data.

0 citationsRead paper