A new benchmark evaluates LLM deductive reasoning in mystery stories by swapping character names and shows accuracy varies with name familiarity.
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
-
WHODUNIT: Evaluation benchmark for culprit detection in mystery stories
A new benchmark evaluates LLM deductive reasoning in mystery stories by swapping character names and shows accuracy varies with name familiarity.