Full 4-benchmark datasets, 6,347 evaluated rows per model, GLM-5.3-Flash judge traces, outputs and scoring logs.
Keshav
ASTRALK
·
AI & ML interests
None yet
Recent Activity
updated a dataset 1 day ago
ASTRALK/aya-baselines-20261111 updated a model 1 day ago
ASTRALK/aya-hallucination-evaluation updated a collection 1 day ago
Aya Safety & Hallucination Benchmarks (Datasets)