Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Conversational Quality Evaluation on TGC 10% human-annotated sample

94.37Factual Correctness

TGC

86.51888.556590.59592.6335Apr 14, 2026
Updated 3mo ago

Evaluation Results

MethodLinks
2026.04
94.3797.8288.2294.6797.11-
2026.04
86.82----87.41