Reasoning-tuned DeepSeek models give only small, inconsistent gains over base models on idiomaticity detection, and supplying definitions from large models can improve smaller models on some datasets.
InProceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 16764–16800, Miami, Florida, USA
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
-
Stands to Reason: Investigating the Effect of Reasoning on Idiomaticity Detection
Reasoning-tuned DeepSeek models give only small, inconsistent gains over base models on idiomaticity detection, and supplying definitions from large models can improve smaller models on some datasets.