Pietro Lesci
pietrolesci
AI & ML interests
I like developing and applying causal methods to study the effect of training choices on models’ behaviour, including memorisation, shortcut learning, and tokenisation.
Recent Activity
updated
a dataset
about 5 hours ago
pietrolesci/pile-validation
updated
a collection
about 5 hours ago
Generalisation-Profiles
updated
a collection
about 5 hours ago
Generalisation-Profiles
Organizations
Collections
8
spaces
1
models
19

pietrolesci/me850M_minipile_bpe32000minipile
Updated
•
54

pietrolesci/me340M-tied_minipile_bpe32000minipile
Updated
•
56

pietrolesci/me57M-tied_minipile_bpe2wp32000minipile
Updated

pietrolesci/me57M-tied_minipile_bpe128000minipile
Updated

pietrolesci/me57M-tied_minipile_wordpiece32000minipile
Updated

pietrolesci/me57M-tied_minipile_bpe8064minipile
Updated

pietrolesci/me57M-tied_minipile_bpe32000minipile
Updated

pietrolesci/tokenisers
Updated

pietrolesci/bert-civilcomments-gradtracking
Updated

pietrolesci/roberta-base_mnli_b9799b8f9b
Updated
datasets
54
pietrolesci/pile-validation
Viewer
•
Updated
•
429k
•
58
pietrolesci/pile-deduped-pythia-tokfreq
Viewer
•
Updated
•
50.1k
pietrolesci/pile-deduped-pythia-preshuffled
Viewer
•
Updated
•
97.6M
•
326
pietrolesci/finewebedu-20B
Viewer
•
Updated
•
40.4M
•
141
pietrolesci/me-minipile-evals
Viewer
•
Updated
•
1.82M
•
52
pietrolesci/minipile
Viewer
•
Updated
•
6.06M
•
544
pietrolesci/opus-5langs-1M
Viewer
•
Updated
•
5M
•
124
pietrolesci/opus-raw
Viewer
•
Updated
•
4.06B
•
2.7k
pietrolesci/pythia-pile-stats
Viewer
•
Updated
•
113M
•
176
pietrolesci/slim-pajama-eval
Viewer
•
Updated
•
1.84M
•
61
•
1