Emotion-o1 uses distillation, supervised fine-tuning, and reinforcement learning to make an 8B LLM adjust its chain-of-thought length per emotion task, improving F1 and reducing reasoning cost.
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
-
Emotion-o1: Adaptive Long Reasoning for Emotion Understanding in LLMs
Emotion-o1 uses distillation, supervised fine-tuning, and reinforcement learning to make an 8B LLM adjust its chain-of-thought length per emotion task, improving F1 and reducing reasoning cost.