Rubric changes shift implicit-commitment labels by 70-83% agreement and make some metrics uninformative, so model rankings only stabilize after a metric-identifiability audit.
EBISU: Benchmarking large language models in japanese finance
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Measurement Risk in Supervised Financial NLP: Rubric and Metric Sensitivity on JF-ICR
Rubric changes shift implicit-commitment labels by 70-83% agreement and make some metrics uninformative, so model rankings only stabilize after a metric-identifiability audit.