← Back to feed

BenchAudit – Auditing LLM benchmarks for leaks, data reuse, and harness gaps

AI · Studies · 2026-11-18 07:41

BenchAudit provides an independent index to evaluate the resilience of AI benchmarks against training data contamination, outdated questions, and evaluation framework vulnerabilities.

Source: benchaudit.org


Auto-generated summary for reference only. RSS