A new expert-principle preference dataset enables an 8B LLM to reach 87% win rate vs GPT-4o on counseling responses through standard preference optimization.
Beyond training objectives: Interpreting reward model divergence in large language models
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
-
Preference Learning Unlocks LLMs' Psycho-Counseling Skills
A new expert-principle preference dataset enables an 8B LLM to reach 87% win rate vs GPT-4o on counseling responses through standard preference optimization.