Sebastian Bordt PRO
sbordt
·
AI & ML interests
None yet
Recent Activity
updated
a collection
10 days ago
martin updated
a collection
10 days ago
train-once-answer-all updated
a collection
10 days ago
train-once-answer-all Organizations
None yet
martin
weight-decay
train-once-answer-all
Modes and datasets for the paper "Train Once, Answer All: Many Pretraining Experiments for the Cost of One", ICLR 2026
forgetting-contamination-benchmark-questions
The 44.000 de-duplicated benchmark questions used in "How Much Can We Forget about Data Contamination?" (ICML 2025)