Influence Distillation selects LLM fine-tuning data by approximating each sample's gradient influence on a target task via landmarks and JVP embeddings, matching or beating RDS+ accuracy at roughly one third the selection cost.
Woodruff, and Michael Wunder
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
-
Efficient Data Selection at Scale via Influence Distillation
Influence Distillation selects LLM fine-tuning data by approximating each sample's gradient influence on a target task via landmarks and JVP embeddings, matching or beating RDS+ accuracy at roughly one third the selection cost.