A new automatic method for creating minimal reasoning-preserving variants of NLI problems shows that 14 models drop 4 to 20 percent in accuracy on those variants.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MERGE: Minimal Expression-Replacement GEneralization Test for Natural Language Inference
A new automatic method for creating minimal reasoning-preserving variants of NLI problems shows that 14 models drop 4 to 20 percent in accuracy on those variants.