Archives · every published entry

A worked example of interrupted-time-series NLP on a small, clustered text corpus when the cultural narrative around it already carries strong priors. Pipeline: pre-registered surface metrics with HAC SEs and Bonferroni, sentence-embedding centroid permutation, TF-IDF discriminative interpretability, album-level cluster permutation, and a placebo time-cut calibration. Substantive result on Drake's full discography (2009–2026): the four pre-registered surface metrics are null; the song-level embedding signal (p ≈ 0.0002) does not survive album-level cluster permutation (p = 0.11 unconditional, 0.35 within rap); against 23 placebo cut dates the May 2024 result sits in the bottom quartile of L2 distances; the same pipeline on a peer artist (J. Cole) at the same cutoff also returns null (p = 0.76). Methodological takeaway: with low post-event sample sizes, song-level permutation can manufacture orders of magnitude of apparent evidence; cluster-level inference, placebo time-cut calibration, and a peer-artist control are what determine whether the signal is real.

A two-period self-signaling model in which persistent passive sensors shift identity utility from narrative-compliance to evidentiary consistency. Sensor precision drives two opposing comparative-statics — Goodhart-style wedge exploitation (P1) and self-attribution-bias attenuation (P3) — predicting cross-domain heterogeneity in fitness, glucose, and screen-time tracking. Keystone empirical test inverts the standard signaling prediction.

A masked-autoencoder-pretrained ViT that tokenizes IV surfaces along financial axes and consumes a stack of past surfaces to produce calibrated distributional forecasts of realized volatility. Evaluates whether the representation improves on HAR-RV-IV and supports a profitable variance-swap rule. Synthetic-data pipeline validation only, real OptionMetrics results pending.

The classical secretary problem gives 1/e ≈ 37%, but the rule answers a question almost no one actually asks. Six simulations covering recall, cardinal payoffs, full information, two-sided choice, learned priors, and a market-timing backtest where the naive no-policy baseline beats every stopping rule.

Korea's June 2023 telemedicine pilot collapsed take-up to 0.16% of NHIS claims despite eligibility rules that explicitly targeted chronic-disease elderly. Using DML and causal forests on linked claims data, I estimate the behavioral residual on always-eligible, always-supplied patients and trace roughly half its variance to habit strength and patient–provider trust.