← ClaudeAtlas

algo-probabilistic-concurrency-and-measurementlisted

Use when working with approximate or concurrent structures, or when measuring performance: Bloom filters, HyperLogLog, count-min sketch and streaming structures, benchmarking and profiling discipline, vector and similarity search (HNSW, IVF, approximate nearest neighbour recall), and concurrent data structures — locks, lock-free and wait-free designs, memory ordering, and the ABA problem.
adammatthewsteinberger/vibey-skills · ★ 1 · AI & Automation · score 72
Install: claude install-skill adammatthewsteinberger/vibey-skills
# Algorithms Deep Dive: Probabilistic Structures, Measurement, and Concurrency > **Part 4 of 5** of the *Algorithms Deep Dive* reference (plugin `algorithms-deep-dive`), covering §12–§14. Sibling skills: `algo-foundations-and-machine-model` (§0–§2), `algo-data-structures` (§3–§6), `algo-core-algorithms` (§7–§11), `algo-reference` (§15–§20). Section numbers are shared across the set; a reference written as §N → `skill` points into that sibling skill. > > **Currency:** Verified August 2026. See §17 → `algo-reference` for the currency snapshot and what goes stale first. > **How to read this.** Reference, not a course, and deliberately complementary to a > theory-of-computation reference: **that** answers "is this solvable and how hard is it"; > **this** answers "which structure, which algorithm, and why is it slower than the > analysis said." > > Three markers: > - **[DURABLE]** — established algorithms, structures, and engineering practice. Most of > this document. > - **[VERSIONED]** — library implementations, hardware behaviour, recent results. > - **[CONTESTED]** — genuine disagreement. > > **⚠️ GOTCHA** boxes mark the places where the textbook answer and the production answer > diverge. > > **The three framings that organize everything below:** > 1. **Data structure choice is algorithm choice.** Most "algorithm problems" in real > systems are solved by picking the right structure and letting the algorithm fall out. > **If your algorithm is complicated, your data l