Pietro Lesci

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.

Organizations

liked a dataset over 1 year ago

Zyphra/Zyda-2

Preview • Updated Aug 6, 2025 • 31.2k • 98

liked a Space over 1 year ago

The Ultra-Scale Playbook

🌌

3.95k

The ultimate guide to training LLM on large GPU Clusters

liked 2 models almost 2 years ago

BEE-spoke-data/smol_llama-101M-GQA

Text Generation • 0.1B • Updated Dec 29, 2025 • 479 • 33

BEE-spoke-data/smol_llama-81M-tied

Text Generation • 81.3M • Updated Dec 29, 2025 • 120 • 10

liked 2 datasets over 2 years ago

akjindal53244/Arithmo-Data

Viewer • Updated Jan 26, 2024 • 549k • 67 • 35

Muennighoff/natural-instructions

Viewer • Updated Dec 23, 2022 • 7.15M • 6.9k • 84

liked a dataset almost 3 years ago

JeanKaddour/minipile

Viewer • Updated Jun 20, 2023 • 1.01M • 2.74k • 150

liked 2 Spaces almost 4 years ago

CodeFormer

🐼

2.4k

Restore and enhance faces in photos with optional upscaling

PromptSource

💩