measurement-validity

Here are 3 public repositories matching this topic...

Pranavsingh431 / Perturbation-Based-Measurements

Code, data, and raw model outputs for "Failure Modes in Perturbation-Based Measurement of Language Model Reliability". evaluation/reproduce.py re-derives every reported number offline, no API key required.

benchmark language-models reproducibility uncertainty-quantification llm-evaluation sycophancy research-artifact measurement-validity

Updated Aug 1, 2026
Python

Nasser-Lim / jamo-bench

Star

JAMO — OCR-mediated Hangul rendering benchmark. Code companion to huggingface.co/datasets/Nasser4963/jamo-gold

benchmark ocr hangul text-to-image zenodo korean-nlp glyph-rendering evaluation-methodology ai-benchmarking measurement-validity

Updated Aug 17, 2026
Python

hermes-labs-ai / precise-records-unstable-meanings

Star

Measurement validity, construct validity, and unsupported claims derived from AI agent telemetry.

reproducible-research telemetry research-paper ai-agents construct-validity llm-evaluation agent-evaluation agent-observability ai-reliability epistemic-engineering hermes-labs measurement-validity

Updated Aug 6, 2026
TeX

Add this topic to your repo

To associate your repository with the measurement-validity topic, visit your repo's landing page and select "manage topics."

Learn more