11 projects
pertrank
Reusable control-geometry reliability for gene-specific perturbation prioritization
evcrl
Gymnasium environment for camera-based electric-vehicle eco-driving with sampled smooth execution
pad-causal
PaD: joint interpretation of single- and double-perturbation evidence
certlim
Finite-evidence certification limits: what any Certify/Reject rule can achieve from the evidence you have.
track-certify
Anytime joint certification of causal graphs and unknown intervention targets: certify-or-refuse sequential experiments with instance-optimal adaptive sampling
sbt-monitor
Ledger-first tools for signed boundary transport accounting, calibrated warning, and deployment-cliff diagnostics
witnesscell
Evidence-gated condition-mean prediction for unmeasured genetic combinations
perturbresidual
Auditable functional-coupling scores from measured combinatorial perturbation panels
prior-release-controller
Auditable relative-Global granularity control for learner-conditioned prior release
biohcc
Probe measurement primitives and the BioHCC calibration construction for biomedical hidden-confounding benchmarks
permscore
Joint causal discovery with unknown intervention targets: recover a DAG and which environment perturbed which node
