SFT-GO retrains LLMs by focusing on the worst-performing group of important or unimportant tokens, yielding modest average benchmark improvements over standard supervised fine-tuning.
Training language models to follow instructions with human feedback
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
SFT-GO: Supervised Fine-Tuning with Group Optimization for Large Language Models
SFT-GO retrains LLMs by focusing on the worst-performing group of important or unimportant tokens, yielding modest average benchmark improvements over standard supervised fine-tuning.