JAMA 2016; 316(22) Replication Study
Release package for SlopShape: Identifying AI-Generated Commercial Web Content (verification artifacts, instrument, prompts, code)
This is the reproduction package of the paper Simple Techniques Work Surprisingly Well for Neural Network Test Prioritization and Active Learning by M.Weiss and P.Tonella, published at ISSTA 2022
CS470 Introduction to AI Project of Team CoreFinder
Independent replication of Oliveira et al. (arXiv:2503.11499) on macro regime-conditioned sector ETF allocation
A robot that does paid work, gets paid, and buys the electricity it needs to keep working. Task market, verification oracle, smart bin station and deployment planner for a field study in the Colombian Andes.
Independent reproduction and external validation of image-based deep learning prognostic models for head and neck cancer (Mateus et al., Sci Rep 2023): exact & statistical reproducibility, external validation on an Indian cohort, and CLAIM + TRIPOD+AI audit. Includes code, data, model weights and logs.
Replication and extension of "Currency Momentum Strategies" (Menkhoff, Sarno, Schmeling & Schrimpf, 2012) over 2000-2024, using Datastream FX data. Tests momentum against transaction costs, carry, Fama-MacBeth regressions, and equity/FX risk factors.
Replication study of finite-time convergent continuous-time optimization flows (q-RGF and RNF) from Romero & Benosman (ICML 2020), implemented in Python.
Replication of a high-precision Fourier analysis method (DFT¹, signal-derivative-based) for sub-bin frequency and amplitude estimation
Validated Korean equity market replication of Frazzini and Pedersen (2014) Betting Against Beta
Independent reproduction and external validation of image-based deep learning prognostic models for head and neck cancer (Mateus et al., Sci Rep 2023): exact & statistical reproducibility, external validation on an Indian cohort, and CLAIM + TRIPOD+AI audit. Includes code, data, model weights and logs.
A pre-registered replication of 12 published trading strategies. Two gates, both reported, plus an errata documenting seven defects found in my own code - one of which had been hiding a conclusion.
Replicating latent-space backdoor leakage and behavioral transfer in LLMs using Pythia-70M.
Independent replication and audit of 'Proving and Improving the Platonic Representation Hypothesis with JEPAs' (Balestriero and LeCun, 2026) on one Mac
Replication and stress-test of The Universal Weight Subspace Hypothesis (arXiv:2512.05117). Four experiments on the paper's own corpus; finds three of four headline claims are not supported at this replication scale.
LLM agents in place of human subjects in Plott & Sunder (1982)'s five insider-information markets, plus three new controls. Engine, JSONL logs, replay viewer.
Replication study: the 70–90% accuracies reported for BTC daily direction prediction vanish under leak-proof evaluation (~52%, the base rate). Companion experiment shows the usual leaks barely matter, but a one-line label-misalignment bug hands a tabular model 97.5% while the CNN shrugs it off.
Simulation-only replication of a Koch–Sierpinski hybrid fractal antenna (Azzouz et al., Fractal Fract. 2026, 10, 47). Ansys HFSS + Python.
A viral claim: 'Trump tweets predict the S&P with 61% accuracy.' I reproduced the 61.3% exactly, then tested whether it actually makes money on Polymarket. Pre-registered 7-gate test. Verdict: KILL — 0/100 cells cleared.
To associate your repository with the replication-study topic, visit your repo's landing page and select "manage topics."