A framework using Taylor expansions and random projections estimates LLM performance under arbitrary training batch orders from one reference run.
If influence functions are the answer, then what is the question?Advances in Neural Information Processing Systems, 35:17953–17967, 2022
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
REJECT 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Estimating the Effects of Sample Training Orders for Large Language Models without Retraining
A framework using Taylor expansions and random projections estimates LLM performance under arbitrary training batch orders from one reference run.