Kyle O'Brien's picture

Kyle O'Brien PRO

Kyle1668

·

https://kyobrien.io

AI & ML interests

Interpretability, model editing, alignment

Recent Activity

updated a model about 17 hours ago

Unlearning/pythia1.5_modernbert_filtered_5percent_wmdp_deep_fry_20x_upsampled

published a model about 17 hours ago

Unlearning/pythia1.5_modernbert_filtered_5percent_wmdp_deep_fry_20x_upsampled

updated a model 4 days ago

Unlearning/pythia1.5_blocklist_filtered_wmdp_lie_o_rewrite_20x_upsampled

View all activity

Organizations

Collections 1

Papers 2

arxiv:2402.08225

arxiv:2304.01373

models 27

Kyle1668/answerdotai-ModernBERT-large_20250111-002259

Text Classification • Updated Jan 14 • 4

Kyle1668/answerdotai-ModernBERT-large_20250111-224237

Text Classification • Updated Jan 12 • 2

Kyle1668/answerdotai-ModernBERT-large_20241230-093521

Text Classification • Updated Dec 30, 2024 • 14

Kyle1668/allenai-scibert_scivocab_uncased_20241230-091934

Text Classification • Updated Dec 30, 2024 • 11

Kyle1668/boss-toxicity-bert-base-uncased

Text Classification • Updated Feb 7, 2024 • 23

Kyle1668/ag-news-t5-large

Text2Text Generation • Updated Feb 7, 2024 • 14

Kyle1668/ag-news-76800-bert-base-uncased

Text Classification • Updated Nov 7, 2023 • 11

Kyle1668/ag-news-38400-bert-base-uncased

Text Classification • Updated Nov 7, 2023 • 14

Kyle1668/ag-news-19200-bert-base-uncased

Text Classification • Updated Nov 7, 2023 • 1.67k

Kyle1668/ag-news-9600-bert-base-uncased

Text Classification • Updated Nov 7, 2023 • 9

datasets 7

Kyle1668/mmlu_auxiliary_train_formatted

Viewer • Updated 21 days ago • 99.8k • 80

Kyle1668/phi_sae_training

Viewer • Updated Sep 7, 2024 • 17.2M • 58

Kyle1668/LLM-TTA-Cached-Rewrites

Viewer • Updated Feb 7, 2024 • 986k • 19

Kyle1668/LLM-TTA-Augmentation-Logs

Viewer • Updated Feb 7, 2024 • 4.43M • 47

Kyle1668/AG-Tweets

Viewer • Updated Nov 29, 2023 • 7.6k • 19

Kyle1668/BOSS-Robustness-Benchmark

Preview • Updated Oct 11, 2023 • 11

Kyle1668/pythia-semantic-memorization-perplexities

Viewer • Updated Sep 19, 2023 • 99.7M • 492