M&S8 named
Models & Statistics
What changes when information arrives late, moves underneath you, or refuses to behave like the model?
Currently: Clock Dataset — What was actually knowable at the moment of a decision?
Tune in- 01
Clock Dataset
What was actually knowable at the moment of a decision?
- 02
Decision Assurance
When does a score stop being enough for a finite-capacity decision?
- 03
Degradation Diagnosis
How much can a model’s trouble be diagnosed before its labels arrive?
- 04
Release Equivalence
Is the system after release still the one you tested?
- 05
Delayed Feedback
Evaluating policies when outcomes arrive late, selectively, and after the policy has changed.
- 06
DeepLOB
How an experiment can manufacture confidence from a market with no signal.
- 07
Spectral Recovery
A controlled look at when correlation-matrix denoising helps—and when it does not.
- 08
Tail Estimator Audit
Known-tail experiments for finding where familiar estimators bend or break.
- 09
Evidence that exists and cannot be retrieved
A successful empty audit query did not establish an empty record — 192 records sat in the log the whole time.
- 10
Systems & Security Core
Range harnesses, evidence ledgers, and provenance tools for difficult systems questions.
- 11
Security Decision Invariance
When a transformed event record still supports the same security conclusion.
- 12
Agentic Systems
A deliberate open shelf for work on autonomous software, tools, and accountable coordination.
- 13
GPU Simulation Engine
GPU-first compute and rendering for typed simulation state.
- 14
Cascade Field Study
A deterministic field for watching a local shock remain bounded—or become a cascade.