Integrating TEE-certified DP training into our Bayesian optimization pipeline for virtual screening will maintain surrogate model sample efficiency (measured by regret) while keeping the underlying proprietary screening dataset cryptographically verifiable to external collaborators.

Adversarial Debate Score

51% survival rate under critique

Supporting Research Papers

• Empowering Chemical Structures with Biological Insights for Scalable Phenotypic Virtual Screening

  Motivation: The scalable identification of bioactive compounds is essential for contemporary drug discovery. This process faces a key trade-off: structural screening offers scalability but lacks biolo...

• Adaptive Methods Are Preferable in High Privacy Settings: An SDE Perspective

  Differential Privacy (DP) is becoming central to large-scale training as privacy regulations tighten. We revisit how DP noise interacts with adaptivity in optimization through the lens of stochastic d...

• Inference-time optimization for experiment-grounded protein ensemble generation

  Protein function relies on dynamic conformational ensembles, yet current generative models like AlphaFold3 often fail to produce ensembles that match experimental data. Recent experiment-guided genera...

• Empirical Bayes prepivoting under group invariance: false discovery rate control and moderated t-tests

  We consider simultaneously testing hypotheses about thousands of units, e.g., genes or proteins, where each unit yields a handful of replicate measurements and we test whether its mean is zero. A wide...

• Prediction-Powered Smoothing and Validation for Disaggregated AI Evaluation

  Evaluating an AI system requires disaggregated assessment, as performance varies across domains such as benchmark task types or conversation types in deployed agents. Exhaustive testing is expensive, ...

Formal Verification

Source