Geometric Differences in FDR Control
lihua_lei_stat · x · 2026-07-19
In a series of conclusions, the author points out that their constructed example defies the common intuition that "negative correlation breaks FDR control." Instead, it is a positively correlated construction; it satisfies the PRDN proposed by @weijie444 but fails to meet the standard PRDS.
The author uses this to demonstrate that the geometric structures of one-sided and two-sided tests differ vastly, meaning they cannot be understood through the same intuitive lens.
Related event: Positive Correlation Counterexample Highlights FDR Control Differences(2 posts)→
More from Research
- Stanford Team Introduces Gigatoken, the World's Fastest Tokenizer — StanfordAILab · 2026-07-22
- Tabul AI launches Metal TreeSHAP to speed up Shapley values on Apple silicon — Scobleizer · 2026-07-22
- Reddit points to OpenAI’s ChatGPT Ads page — EcstaticAsparagus509 · 2026-07-22
- Open-source runtime lets each repo define its own AI code reviewer — ibabufrik · 2026-07-22
- DeepSWE: A New Benchmark for Evaluating AI Coding Agents on Real GitHub Issues — pmz · 2026-07-22
- A Rust space-economy sim runs hundreds of autonomous ships, built with Claude — kalcode · 2026-07-22