Vault Index

Routing Summary

A second brain for research, clippings, and knowledge management. Contains 130+ ingested notes organized by topic.

• Need Bayesian statistics (BDA3, Statistical Rethinking, Workflow)? → Bayesian Statistics
• Need econometrics and causal inference (MHE)? → Econometrics
• Need research methodology (forking paths, power analysis, longitudinal causal inference)? → Research Methodology
• Need agent-based modeling (consumer behavior, WOM, diffusion)? → Agent-Based Modeling
• Need web clippings and saved articles? → Clippings
• Need market response models (functional forms, carryover, VAR, advertising/price elasticities)? → Market Response Models
• Need category theory (functors, Yoneda lemma, limits, adjoint functor theorems)? → Category Theory
• Need causal discovery / DAG structure learning from data (NOTEARS, continuous optimization)? → Causal Discovery
• Need machine learning / AI (transformers, LLM scaling laws, chain-of-thought, RAG, agents, RLHF, conformal prediction, probabilistic forecasting)? → Machine Learning and AI
• Need causal machine learning (DML, causal forests) or synthetic DiD / event studies? → Causal Machine Learning, Synthetic Difference-in-Differences
• Need online A/B testing statistics or customer lifetime value models? → Online Experimentation, Customer Lifetime Value

Areas

Area	Notes	Domain
Research	595	Applied statistics, econometrics, causal inference, causal discovery, Bayesian experimental design, probabilistic numerics, agent-based modeling, market response models, category theory from textbooks and papers, machine learning and AI
Clippings	13	Web articles and saved content (raw source material)

Topic Map

Cross-cutting topics that span multiple folders:

• Bayesian Statistics: BDA3, Workflow, Hierarchical Models, MCMC, Model Checking, Model Comparison
• Causal Inference: Selection Problem, Experiments, CIA, IV, DD, Activity Bias, Counterfactual Inference, BART Causal, Potential Outcomes Framework, Causal Estimands, Propensity Score in Bayesian CI, Synthetic Control, SC Bias Theory, SC Requirements, GSC, Metalearners, CausalImpact BSTS, LLM-BN Elicitation, Cause-Precondition-Effect, DAG Summarization, CaGReS, s-Separation, Code Causal Prompts, LLM Causal Tasks, Conditional Fine-tuning, Dynamic Treatment Regimes, Q-learning, A-learning and Robustness
• Econometrics: CEF, OVB, Quantile Regression, Standard Errors, Sim-Based Est., MSM, SMM Weighting, SMM Code, Indirect Inference, EMM, SMM Copulas, SME (Duffie-Singleton), SME Consistency, SME Asymptotic Distribution, Plausible GMM, Plausibility Characteristic, Plausibility-Adjusted Weighting
• Research Methodology: Forking Paths, Researcher Degrees of Freedom, Bonferroni, Power Analysis, Survival Analysis, Delayed Feedback, Censored Bandits
• Advanced Models: Nonparametric, HSGP, Spatial, Copulas, Dependence Measures, Networks, SEM
• Likelihood-Free / Simulation-Based Inference: Synthetic Likelihood, Synthetic Likelihood Construction, Chaos and Phase-Insensitive Statistics, ABC, MSM, SME, Indirect Inference
• Agent-Based Modeling: ABM Foundations, Consumer Utility, CUBES Behavioral Model, WOM, Diffusion, GA Calibration, HM+ABC, History Matching, ABC for ABMs
• Longitudinal Methods: Between Overview, FE Model, DPM, Longitudinal Estimands
• Market Response Models: MRM Overview, Functional Forms, MNL, ADL, Reaction Functions, Estimation, ARIMA, Transfer Function, Cointegration, Advertising Generalizations, Price Generalizations, Geo-Experiment Methodology, Matched Markets
• Machine Learning and AI: Transformers and LLM Foundations - Overview, Neural Scaling Laws, In-Context Learning and Few-Shot Prompting, LLM Reasoning, Retrieval and Agents - Overview, Retrieval-Augmented Generation (RAG), Tool Use and the Agent Loop, Conformal Prediction - Overview, Probabilistic Forecasting - Overview; bridges: Causal Machine Learning - Overview, Variational Inference - Overview, Neural Simulation-Based Inference - Overview, LLM-Powered Agents - Overview
• Experimentation and Customer Analytics: Online Experimentation - Overview, CUPED and Regression-Adjusted Variance Reduction, Confidence Sequences, Synthetic Difference-in-Differences - Overview, Honest DiD - Sensitivity to Parallel Trends Violations, Customer Lifetime Value - Overview, BG-NBD Model
• User-Level Ad Experiments: User-Level Ad Experiments - Overview, Intent-to-Treat, PSA and Ghost Ad Designs, From ITT to Treatment-on-the-Treated in Ad Experiments, The Unfavorable Economics of Ad Experiments - Power and Signal-to-Noise, Experimental Benchmarks for Observational Ad Measurement, User-Level vs Geo-Level Experiments - When to Use Which
• Bayesian Experimental Design: BED Overview, Lindley 1956, EIG, NMC, Variational EIG Estimators, VNMC, Unified Gradient BOED, ACE, PCE, Adaptive Design, Deep Adaptive Design, Modern BED Review, Thompson Sampling, Regret Bounds
• Probabilistic Numerics: PN Overview, Computation as Inference, Numerical Agent, Gaussian Algebra, GP Regression, RTS, Bayesian Quadrature, Classical Quadrature as Inference, Probabilistic Linear Solvers, CG as Inference, Bayesian Optimisation, Acquisition Functions, ODE Filters, ODE Filter Convergence
• Causal Discovery: NOTEARS Overview, DAG Structure Learning, Smooth Acyclicity h(W), NOTEARS Algorithm, NOTEARS Experiments
• Category Theory: BCT Overview, Categories, Functors, Natural Transformations, Adjoint Functors, Yoneda Lemma, Limits, RAPL Theorem, SAFT

Recent Ingestions

• 2026-09-18: Consolidation Q&A batch + user-level testing. Answered 20 cross-cluster consolidation questions from the vault’s own notes (≈60k words, 437 source-note citations, 431 used_by backlinks) — see Questions and Answers (now 29). Highlights: Q - Comparing Geo-Test Estimators from TBR to Synthetic DiD, Q - Using Experiment Results as Priors in a Bayesian MMM, Q - Covariate Adjustment for Precision vs Identification, Q - The Common Structure of Doubly-Robust Estimators, Q - Does Peeking Matter for a Bayesian, Q - Choosing a Simulation-Based Inference Method for ABM Calibration. New cluster User-Level Ad Experiments (9 notes from Johnson–Lewis–Nubbemeyer ghost ads, Lewis–Rao, Gordon et al. 2019/2023, Lin–Misra, Johnson–Lewis–Reiley). A Q&A also caught and fixed a formula error in Optimal Marketing Decisions and Forecasting (multiplicative-model optimum is $\alpha m=\alpha /\beta$, not $\alpha /m$).

• 2026-09-18: Graph cleanup and de-duplication. Rebuilt the vault graph (now 1 connected component, 0 orphans) and removed every ambiguous link target: merged the Theoretical Physics/ folder into Physics (duplicate overviews combined into Quantum Field Theory - Overview and Gauge Theory - Overview; 5 notes moved); merged the two Bayesian IPW notes into Bayesian Inverse Probability Weighting; merged the horseshoe stub into Horseshoe and Regularized Horseshoe Priors; retitled the loose PSM synthesis note to Propensity Score Matching - Balancing Theorem and Failure Modes and moved it with its two companions into Propensity Score Matching; removed 52 aliases that collided with another note’s title or alias; added 25 missing links suggested by shared-neighbour analysis.

• 2026-09-18: Gap-fill research run (AI + applied-statistics use cases) — audited the vault for gaps, then researched and ingested 49 primary papers into 85 new notes across 11 new clusters (+12 indexes), with 98 See Also backlinks and 115 used_by entries added to 89 existing notes.

  • New topic Machine Learning and AI (29 notes): Transformers and LLM Foundations (Vaswani 2017, Kaplan 2020, Hoffmann 2022, Brown 2020), LLM Reasoning, Retrieval and Agents (Wei 2022, Lewis 2020, Yao 2022, Ouyang 2022), Conformal Prediction (Angelopoulos & Bates 2021, Romano 2019, Tibshirani 2019, Lei & Candès 2020), Probabilistic Forecasting (Gneiting & Raftery 2007, Salinas 2017, Ansari 2024, Wickramasuriya 2019).
  • AI bridges into existing topics: LLM-Powered Agents (8; generative agents, silicon samples, homo silicus), Neural Simulation-Based Inference (8; NPE/NLE/NRE, flows, ABM calibration), Variational Inference (8; Dream gap #33), Causal Machine Learning (9; Dream gaps #40, #46).
  • Applied-statistics use cases: Synthetic Difference-in-Differences (8; Dream gaps #45, #47, incl. SDID for Geo Experiments and Marketing Panels), Online Experimentation (8; CUPED, mSPRT, confidence sequences, SRM, switchbacks), Customer Lifetime Value (7; Pareto-NBD, BG-NBD, Gamma-Gamma, sBG).
  • Remaining and newly identified gaps are tracked in Dream: Research Gaps (new section “AI-Era and Applied Gaps”).

• 2026-08-18: Ingested the textbook Gelman, Vehtari & McElreath, Bayesian Workflow (2026; 550 pp., 31 chapters + 2 appendices) into the existing Bayesian Workflow folder — 79 content notes across 7 sub-folders + a book overview + 7 indexes + 3 .base views, expanding the 7 notes previously derived from the 2020 arXiv paper (kept in place, each given an expanded_by frontmatter field and a callout pointing to its book counterpart). Start at Bayesian Workflow Book - Overview.

  • Foundations (8, Ch. 1–4): Why Bayes - Benefits, Costs, and Borders, Varieties of Bayesian Theory, There Is No Safe Haven, From Inference to Data Analysis to Workflow (Figure 2.1 transcribed as mermaid — the book’s routing table), Four Modeling Scenarios, Computational Tools and Probabilistic Programming, Bioassay - A First Probabilistic Program, Multiple-Choice Exam - A Full Workflow Walkthrough.
  • Building Models (19, Ch. 5–7): model specification, every prior-choice rule (Prior Distributions, Constructing Priors for Effect Sizes, Joint Priors and Covariance Matrices), Prior Predictive Checking, Tail Behavior and Prior-Likelihood Conflict, Generative and Partially Generative Models, Designing Simulated-Data Experiments, Poststratification, Causal Inference as Generalization, From Inference to Decision.
  • Evaluating and Comparing (16, Ch. 8–10): Posterior Predictive Checking, Cross Validation Checking (PSIS-LOO, Pareto $\hat{k}$, p_loo, LOO-PIT-ECDF), Influence of Likelihood and Prior (power-scaling), Model Selection Using Predictive Performance, Model Selection and Overfitting, Stacking and Predictive Model Averaging, Model Expansion - Predictive Consistency and Coherence, Simulated-Data Experimentation as Virtual Replication.
  • Computational Workflow (16, Ch. 11–13, 15): The Typical Set and the Log Posterior Density, Chains, Iterations, and Effective Sample Size, Effective Sample Size and Monte Carlo Standard Error, Fit Fast, Fail Fast, Failure Modes and Steps Forward (funnels, aliasing, label switching, the folk theorem), What to Do About Convergence Problems, Variational Inference and Pathfinder, Statistical Modeling as Software Development.
  • Simulation-Based Calibration (1, Ch. 14): SBC in the Workflow — complements, rather than supersedes, the 6 existing Talts et al. (2018) theory notes.
  • Case Studies (16, Ch. 16–31, indexed by symptom rather than domain): movie ratings, sleep study, clinical trial, coronavirus testing, classification competition, dogs, black cat adoptions, World Cup football, roaches, golf putting, animal movement, birthdays (Hilbert-space GP, two acknowledged failures), student grades (piranha principle, R2D2, projpred), no vehicles in the park (sum_to_zero_vector; centered beats non-centered), planetary motion (Pathfinder initialization), SBC in model development (five real bugs, the Fano-factor filter).
  • Appendices (2): Statistical and Computational Workflow for Bayesians and Non-Bayesians (a full Bayesian→non-Bayesian translation table) and How to Get the Most Out of Bayesian Data Analysis (a chapter-by-chapter BDA3 reading guide with the authors’ own retractions).
  • Also: Workflow, Bayesian Statistics, and Research indexes updated; this index updated.

• 2026-07-03: Gap-filling ingestion — 5 papers closing three self-identified vault gaps (15 content notes + 3 leaf indexes):

  • Bandits/Thompson sampling (the vault’s previously-flagged “biggest gap”) → new sub-topic Multi-Armed Bandits and Thompson Sampling under Bayesian Experimental Design: Russo, Van Roy, Kazerouni, Osband & Wen (2018), A Tutorial on Thompson Sampling (arXiv:1707.02038) — 6 notes (Overview, Bernoulli Bandit and Thompson Sampling Algorithm, UCB and Greedy Algorithms for Bandits, Regret Bounds for Thompson Sampling, Contextual and Linear Bandits, Approximate Thompson Sampling and Practical Extensions). Resolves the gap named in Q - BED vs Bayesian Optimization vs Bandits for Media Experimentation.
  • Geo-experiment methodology (matched-market design / GeoLift / time-based regression) → new sub-topic Geo-Experiment Methodology under Market Response Models: Vaver & Koehler (2011), Measuring Ad Effectiveness Using Geo Experiments, and Kerman, Wang & Vaver (2017), Estimating Ad Effectiveness using Geo Experiments in a Time-Based Regression Framework (Google) — 4 notes (Overview, Geo-Experiment Design and Power Analysis, Time-Based Regression Estimator for Geo Experiments, TBR Design Sensitivity and the Stationarity Assumption). Resolves the gap named in Q - Encoding a Geo-Holdout as a Bayesian Experimental Design and Computing Its EIG; cross-links GeoLift’s Augmented Synthetic Control machinery to the vault’s existing Synthetic Control/Generalized Synthetic Control Method notes.
  • Experimental design under delayed/censored feedback → new sub-topic Delayed and Censored Feedback under Research Methodology/Experimental Design: Chapelle (2014), Modeling Delayed Feedback in Display Advertising (KDD), and Vernade, Cappé & Perchet (2017), Stochastic Bandit Models for Delayed Conversions (arXiv:1706.09186) — 5 notes (Overview, Delayed Feedback Model for Conversion Prediction, EM and Gradient Optimization for the Delayed Feedback Model, Bandit Models with Delayed and Censored Feedback, plus index). Extends the existing Survival Analysis note’s censoring concept into the online sequential-decision setting.
  • Also: Bayesian Experimental Design, Market Response Models, and Research Methodology parent indexes updated; this index updated.

• 2026-07-01: Ingested the textbook Hennig, Osborne & Kersting, Probabilistic Numerics: Computation as Machine Learning (Cambridge University Press, 2022; 412 pp.) into new top-level topic Probabilistic Numerics — 36 content notes across 5 sub-topics + overview + 5 indexes + 3 .base views. Foundations (7: Computation as Probabilistic Inference, The Numerical Agent, Gaussian Distributions and Algebra, Gaussian Process Regression, Gauss-Markov Processes and SDEs, Bayesian Filtering and Smoothing, Hierarchical Inference in Gaussian Models); Integration (7: Bayesian Quadrature, Kernel Quadrature and Kernel Means, Classical Quadrature as Inference, Convergence and Priors in Bayesian Quadrature, Active Bayesian Quadrature and Bayesian Monte Carlo, Lessons from Integration, The Integration Problem); Linear Algebra (7: Probabilistic Linear Solvers - Algorithmic Scaffold, Conjugate Gradients as Probabilistic Inference, Classic Linear Solvers - A Review, Gaussian Priors over Matrices and the Symmetric Kronecker Product, Computational Constraints on Probabilistic Solvers, Uncertainty Calibration for Linear Solvers, The Linear Algebra Problem and Evaluation Strategies); Optimisation (8: The Local Optimisation Problem, Probabilistic Step-Size Selection and Line Searches, First- and Second-Order Optimisation Methods, The Global Optimisation Problem, Bayesian Optimisation, Value Loss and Entropy Search, Acquisition Functions, Further Topics in Global Optimisation); Differential Equations (6: Solving ODEs as Inference, Classical ODE Solvers as Regression, ODE Filters and Smoothers, Theory of ODE Filters and Smoothers, Perturbative ODE Solvers, Further Topics in ODE Solvers). Unifying thesis throughout: classical solvers (trapezoid/Gauss quadrature, Conjugate Gradients, BFGS, Runge–Kutta) are the posterior mean of a specific Gaussian inference procedure; a solver is a decision-making agent. Research/_Index.md and this index updated.

• 2026-06-27: Ingested 4 papers on Bayesian experimental design into new top-level topic Bayesian Experimental Design — 21 notes across 4 sub-topics. Foundations (Lindley’s Information Measure — the founding 1956 paper: average-information measure, Theorems 1–9, design rule, determinant criterion; Expected Information Gain, Nested Estimation and Nested Monte Carlo, Sequential and Adaptive BED); Variational EIG Estimators (Foster et al. 2019, NeurIPS — Overview, posterior/Barber–Agakov, marginal, VNMC, implicit-likelihood, convergence/selection); Gradient-Based Unified BOED (Foster et al. 2020, AISTATS — Overview, ACE, PCE, likelihood-free ACE & gradients, 100–400-D applications); Modern BED Review (Rainforth et al. 2023, Statistical Science — objectives vs Fisher information, computational revolution/MLMC, gradient optimization, deep adaptive design, open challenges). Also: 5 indexes + 3 .base views.

• 2026-06-27: Ingested Statistical Inference for Noisy Nonlinear Ecological Dynamic Systems (Wood, Nature 466, 2010) into new sub-topic Synthetic Likelihood — 4 notes: Overview, Chaos and Phase-Insensitive Statistics (Ricker map, likelihood collapse, summary-statistic design), Synthetic Likelihood Construction (MVN $l_s(\theta )$, MCMC, MLE via quadratic regression, ${\chi }^2$ diagnostic), Nicholson’s Blowfly Application (Gurney–Nisbet model, full-vs-demographic AIC, limit-cycles conclusion). The foundational likelihood-free / simulation-based inference method, bridging the vault’s ABC and MSM/SME clusters. Also: 1 leaf index + 3 .base views.

• 2026-06-27: Vault repair — fixed 20 broken/renamed wikilinks (Bayesian Propensity Scores and IPW ×11, Mostly Harmless Econometrics ×4, Bayesian Inference, QM/QFT overviews, Experimental Design for ABMs); created Horseshoe and Regularized Horseshoe Priors stub (3 inbound refs); repaired frontmatter on 3 book-overviews + 1 used_by asymmetry; standardized type/book-overview→type/overview. Processed orphaned q- and a- learning.pdf (Schulte, Tsiatis, Laber & Davidian 2014) into new sub-topic Dynamic Treatment Regimes — 5 notes: Overview, Dynamic Treatment Regimes Framework (potential outcomes, optimal regime, identification assumptions), Optimal Regime via Dynamic Programming (Q/value functions, backward induction, midstream invariance), Q-learning (backward regression), A-learning and Robustness (contrast functions, g-estimation, double robustness, simulations). Also: 1 leaf index + 3 .base views.

• 2026-06-27: Ingested Simulated Moments Estimation of Markov Models of Asset Prices (Duffie & Singleton, Econometrica 61(4), 1993) into existing Simulation-Based Estimation — 7 notes: SME Overview (2 simulation challenges), Duffie-Singleton Asset-Pricing Model (stochastic-growth model w/ taste shocks, Eqs. 2.1–2.6), Simulated Moments Estimator Definition ($H,f$ primitives, $G_T$, SME — Eqs. 3.1–3.5), Geometric Ergodicity and Uniform LLN (Condition B, Lemmas 1–2), SME Consistency (Thm 1 weak; AUC/$L^2$-UC conditions; Thms 2–3 strong), SME Asymptotic Distribution (Thm 4, Cor 3.1, $(1+\tau )$ inflation), SME Extensions and Applications (β-dependent moments, calculated-vs-simulated efficiency, option pricing). Foundational time-series theory underpinning the folder’s MSM/SMM notes. Also: sub-topic + parent indexes updated, 3 .base views.

• 2026-06-27: Ingested Plausible GMM: A Quasi-Bayesian Approach (Chernozhukov, Hansen, Kong & Wang, 2026; arXiv:2507.00555, econ.EM) into new sub-topic Plausible GMM — 5 notes: Overview (5 contributions, literature map), Plausible Moment Restriction Model ($m({\theta }_{\ast })={\mu }_{\ast }$, dogmatic vs. plausible prior, plausible-IV example), Quasi-Bayes for Plausible Moment Restrictions (CU-GMM criterion $Q_T$, quasi-posterior $p_T$, optimal decisions — Eqs. 1–3), Gaussian Local Prior Approximation (local prior $\mu \sim \mathcal{N}({\mu }_0,\Lambda /T)$, plausibility-adjusted weighting $A_{\theta }$, “no free lunch” — Eq. 5), Plausible GMM - Institutions and GDP Application (Acemoglu-Johnson-Robinson 2001 IV revisit, Fig. 1 prior-sensitivity). Also: 1 leaf index + 3 .base views. Gap: §4 Bernstein–von Mises theorems and the 401(k) IV-quantile application are in the unincluded Supplemental Appendix.

• 2026-06-17: Ingested 4 research papers (24 notes total) across four topics:

  • DiD → Difference-in-Differences: Callaway & Sant’Anna (2020), Difference-in-Differences with Multiple Time Periods — 6 notes (Overview, Group-Time Average Treatment Effects, Identifying Assumptions for Staggered DiD, Doubly-Robust Estimands for ATT(g,t) (Theorem 1), Aggregating Group-Time Effects, Simultaneous Inference via Multiplier Bootstrap (Theorems 2–3) + minimum-wage application).
  • Factor copulas → new Dependence Modeling: Oh & Patton (2012), Modelling Dependence in High Dimensions with Factor Copulas — 6 notes (Overview, Factor Copula Construction, Tail Dependence in Factor Copulas (Props 1–3, EVT), Multi-Factor and Block Dependence Structures, SMM Estimation of Factor Copulas, Factor Copula Application - S&P 100 and Systemic Risk).
  • SBC → Bayesian Workflow: Talts et al. (2018), Validating Bayesian Inference Algorithms with Simulation-Based Calibration — 6 notes (Overview, Data-Averaged Posterior Self-Consistency, Rank Statistics and Uniformity (Theorem 1), The SBC Algorithm, Interpreting SBC Histograms, SBC Case Studies).
  • Bayesian MMM → new Bayesian Media Mix Modeling: Jin et al. (Google, 2017), Bayesian Methods for Media Mix Modeling with Carryover and Shape Effects — 6 notes (Overview, Carryover (Adstock) Functional Forms, Shape (Saturation) Effects, Bayesian Estimation and Priors for MMM, ROAS, mROAS, and Optimal Media Mix, MMM Model Selection and Application).
  • Also: 3 new leaf indexes + 1 merged (Workflow), 2 new parent sub-topics, parent/root indexes updated, 3 .base dynamic views.

• 2026-06-17: Ingested DAGs with NO TEARS: Continuous Optimization for Structure Learning (Zheng, Aragam, Ravikumar & Xing, NeurIPS 2018; arXiv:1803.01422) into new topic Causal Discovery — 5 content notes: Overview (4 contributions, undirected-GM analogy), DAG Structure Learning Problem (linear SEM, LS score, Programs 3 & 4, NP-hardness, prior-methods landscape), Smooth Acyclicity (Props 1–2, Theorem 1 $h(W)=\mathrm{tr}{e}^{W\circ W}-d$ + gradient, sign-cancellation example), NOTEARS Algorithm (ECP, augmented Lagrangian, L-BFGS/PQN, thresholding, Algorithm 1), Experiments (vs FGS, SHD/FDR, GOBNILP global-optimum, Sachs data). Also: 1 index file + 3 .base dynamic views.

• 2026-05-08: Ingested Basic Category Theory (Leinster, Cambridge 2014; arXiv 1612.09375v2, 191 pp.) into Category Theory — 21 content notes across 6 sub-folders: Foundations (4: categories, functors, natural transformations, functor categories), Adjunctions (3: adjoint functors, units/counits, initial objects), Representables (3: representable functors, Yoneda lemma, Yoneda embedding), Limits and Colimits (5: products/equalizers, pullbacks, general limits, colimits, functors and limits), Synthesis (5: limits via representables, presheaf categories, adjoints and limits, GAFT/SAFT, CCC), Universal Properties (1: introduction). Also: 1 overview note, 7 index files, 3 .base dynamic views.

• 2026-04-11: Ingested 2 papers: (1) McCulloch et al. (2022) into Calibration Methods — 5 notes: HM-ABC Framework overview, History Matching (implausibility score, waves, LHS), Approximate Bayesian Computation (rejection sampling, HM-informed prior, epsilon threshold), Uncertainty Quantification (4 sources: parameter/model discrepancy/ensemble variance/observation), Case Studies (SugarScape, territorial birds, RISC Scottish farms). (2) Rohrer & Murayama (2023) into Research Methodology — 5 notes: Within/Between Persons overview (3 main claims), Within/Between Causal Inference (ATE proof, time-varying confounders), Fixed-Effects Model (Box 1 DAG, assumptions, limitations), Cross-Lagged and Dynamic Panel Models (CLPM, DPM/RI-CLPM, comparison table), Estimands in Longitudinal Research (theoretical estimand definition, 5-step workflow, consistency in psychology).

• 2026-04-11: Ingested Hanssens, Parsons & Schultz (2001) into Market Response Models — 25 notes across 5 subfolders: Introduction (3 notes: MRM overview, management tasks, data/variables), Static Response Models (4 notes: 10 functional forms with full LaTeX + elasticities, MCI/MNL share models, aggregation bias), Dynamic Response Models (4 notes: Koyck/ADL/PDL carryover, reaction functions, S-shape/hysteresis), Estimation and Testing (4 notes: OLS/GLS/SUR/2SLS/Bayes HB, RESET/specification errors, flexible forms, model selection), Time Series Analysis (4 notes: ARIMA, transfer functions, VAR/cointegration/ECM, Granger causality), Empirical Findings (5 notes: generalizations framework, advertising elasticity ≈ 0.10 / duration 6-9 months, price elasticity ≈ −2.5, Dorfman-Steiner optimization, implementation). Also: 8 index files created + Research/_Index.md and _Vault_Index.md updated.

• 2026-04-11: Ingested Evans (2024) Ch. 19 SMM tutorial into Extensions — 2 new notes: SMM Weighting Matrix and Inference (weighting strategies, two-step Ω̂, Newey-West, Σ̂ via Jacobian) + SMM Python Implementation (scipy workflow, L-BFGS-B eps fix, indirect inference pattern). Also: graph analysis moved Dependence Measures for Copulas from Advanced Models to Extensions, and added 14 missing wikilinks across vault.

• 2026-04-11: Ingested 2 papers into Extensions — 10 notes on simulation-based estimation: MSM, indirect inference, EMM (Liesenfeld & Breitung 1998) + SMM for copulas with asymptotic theory, specification testing, Monte Carlo study, financial application (Oh & Patton 2011). Also 1 note on dependence measures in Advanced Models.

• 2026-04-10: Ingested 3 ABM papers into Agent-Based Modeling — 29 notes across 5 sub-topics: Foundations (5), Consumer Behavior (8), Social Dynamics (6), Calibration and Validation (5), Applications (5). Sources: Bonabeau 2002 (PNAS), Karakaya et al. 2011, Ben Said et al. 2002

• 2026-04-10: Ingested 2 papers into Causal Inference — 12 notes: Liu 2025 (6 notes, Knowledge Elicitation — code prompts for LLM causal reasoning) + Zeng 2025 (6 notes, Foundations — causal DAG summarization, CaGReS algorithm, s-separation, do-calculus soundness)

• 2026-04-10: Ingested 4 papers into Causal Inference — 24 notes across 3 new sub-folders: Knowledge Elicitation (Yamashita 2020 interactive NLP + Shaposhnyk 2025 LLM-for-BN), Treatment Effect Estimation (Künzel 2019 S/T/X-learner metalearners), Time Series Causal Inference (Brodersen 2015 CausalImpact BSTS)

• 2026-04-10: Ingested Xu (2017) into Identification Strategies — 2 notes: Overview + Generalized Synthetic Control Method (IFE model, 3-step estimator, LOO cross-validation, parametric bootstrap, EDR voter turnout application)

• 2026-04-10: Ingested Abadie (2021) into Identification Strategies — 5 notes: Overview, Bias Theory (linear factor model, bias bound, sparsity), Inference & Diagnostics (RMSPE ratio, permutation test, backdating), Requirements (5 contextual + 3 data conditions), Extensions (penalized SC, bias correction, elastic net, matrix completion)

• 2026-04-10: Ingested Li, Ding & Mealli (2022) into Causal Inference — 10 notes covering potential outcomes, Bayesian CI structure, BART/BCF/GP outcome models, propensity score strategies, E-value sensitivity analysis, IV/principal stratification, g-computation

• 2026-04-09: Ingested 9 PyMC tutorials into Research — HSGP, copulas, CFA/SEM, spatial BYM, social networks, causal BART, missing data, counterfactual inference, moderation, Bayesian DiD

• 2026-04-08: Ingested 2 PyMC tutorials — discrete choice models, factor analysis

• 2026-04-08: Ingested 5 textbooks/papers — BDA3, Bayesian Workflow, Statistical Rethinking, Garden of Forking Paths, Activity Bias, Mostly Harmless Econometrics