01 Home02 Research03 Projects04 Notes05 About06 Lab
Home/Notes

A notebook in public,
with the margins showing.

Short research notes on assumptions, baselines and what a result can support. Working notes, not peer-reviewed papers.

A negative control has to be able to fail

Causal inferenceAug 2026 · Research in progress

A placebo test condemned a mediation result, until a positive control showed it condemned everything.

Survivors make every universe look better

Quantitative researchJul 2026 · Working note

Replaying real index membership removed about a third of an equal-weight universe’s return, yet left the comparisons intact.

Make it exact before making it fast

Financial MLJul 2026 · Working note

A bounded KV cache sped up Kronos inference 11.7× with bit-identical output, inside a boundary it has to respect.

Before trusting a backtest

Quantitative researchJul 2026 · Working note

Check what was knowable at each decision, then compare against a simple baseline.

Dates give the month of the underlying analysis.