Pinned Loading

1. reclaim-eval reclaim-eval Public

   Reclaim Evaluation: A lossy memory is worse than an empty one. A judge-free benchmark for whether a compressed memory keeps a language model correctable: the brittle memory failure.

   Python 1

2. factwash factwash Public

   Catch AI rewrites that wash hearsay into fact. A write-time gate for AI memory: deterministic checks, evidence-bearing flags, no LLM judge.

   Python 3

3. dinostomp dinostomp Public

   A verification layer for AI evaluations. Checks the instrument, not just the score: data, scorer, runs, numbers, claims, and itself.

   Python 1

4. ecommerce-case-study ecommerce-case-study Public

   Data analyst case study

   HTML

5. movielens-case-study movielens-case-study Public

   Data analyst case study

   HTML

6. acnh-villager-audit acnh-villager-audit Public

   Audit-first analysis of the ACNH catalog: one case mismatch breaks the join and manufactures a false finding

   HTML