Pietro Lesci PRO

pietrolesci

https://pietrolesci.github.io/

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 11 hours ago

pietrolesci/pile-deduped

updated a dataset 6 days ago

pietrolesci/pythia-deduped-memorisation-profiles

updated a dataset 6 days ago

pietrolesci/pile-validation

View all activity

Organizations

pietrolesci's activity

New activity in JeanKaddour/minipile 6 months ago

Domain and provenance annotation

#1 opened over 1 year ago by

haukur

New activity in HuggingFaceTB/SmolLM-135M 8 months ago

Trapezoidal scheduler with cooldown phase

#4 opened 9 months ago by

maveriq

New activity in EleutherAI/pythia-160m 12 months ago

Tokenizer `merges.txt` files

#5 opened 12 months ago by

pietrolesci

New activity in EleutherAI/pile-deduped-pythia-preshuffled over 1 year ago

Sequence "packing" logic

#2 opened over 1 year ago by

pietrolesci

Pad-only sequences from mmap'ed dataset after a certain index

#1 opened over 1 year ago by

pietrolesci

New activity in EleutherAI/pile-duped-pythia-random-sampled over 1 year ago

Add full sequences (beyond the first 64 tokens)

#1 opened over 1 year ago by

pietrolesci

Add full sequences (beyond the first 64 tokens)

#1 opened over 1 year ago by

pietrolesci

New activity in JeanKaddour/minipile over 1 year ago

Domain and provenance annotation

#1 opened over 1 year ago by

haukur