Maria Sofia Bucarelli
@mariasofiab.bsky.social
👩🎓Postdoc Sapienza, Rome — PhD at Sapienza, Rome
Previously Intern at Amazon Search, Lux — Visiting at Cambridge
Previously Intern at Amazon Search, Lux — Visiting at Cambridge
Reposted by Maria Sofia Bucarelli
Will present this at #CVPR ✈️ See you in Nashville 🇺🇸!
Kudos to the team 👏
Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
Kudos to the team 👏
Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
📢Prepend “Singular” to “Task Vectors” and get +15% average accuracy for free!
1. Perform a low-rank approximation of layer-wise task vectors.
2. Minimize task interference by orthogonalizing inter-task singular vectors.
🧵(1/6)
1. Perform a low-rank approximation of layer-wise task vectors.
2. Minimize task interference by orthogonalizing inter-task singular vectors.
🧵(1/6)
March 11, 2025 at 8:02 AM
Will present this at #CVPR ✈️ See you in Nashville 🇺🇸!
Kudos to the team 👏
Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
Kudos to the team 👏
Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
🚀Call for Submissions – @ai4coolerplanet.bsky.social
Excited to announce the AI for a Cooler Planet workshop @ IJCNN 2025 in Rome
📅Submission Deadline: March 20, 2025
Topics include:
🌱 Climate Predictability, Physics-Informed ML, Weather Forecasting & more!
🔗 sites.google.com/view/ai4cool...
Excited to announce the AI for a Cooler Planet workshop @ IJCNN 2025 in Rome
📅Submission Deadline: March 20, 2025
Topics include:
🌱 Climate Predictability, Physics-Informed ML, Weather Forecasting & more!
🔗 sites.google.com/view/ai4cool...
February 20, 2025 at 3:25 PM
🚀Call for Submissions – @ai4coolerplanet.bsky.social
Excited to announce the AI for a Cooler Planet workshop @ IJCNN 2025 in Rome
📅Submission Deadline: March 20, 2025
Topics include:
🌱 Climate Predictability, Physics-Informed ML, Weather Forecasting & more!
🔗 sites.google.com/view/ai4cool...
Excited to announce the AI for a Cooler Planet workshop @ IJCNN 2025 in Rome
📅Submission Deadline: March 20, 2025
Topics include:
🌱 Climate Predictability, Physics-Informed ML, Weather Forecasting & more!
🔗 sites.google.com/view/ai4cool...
Reposted by Maria Sofia Bucarelli
"Task Singular Vectors: Reducing Task Interference in Model Merging" by Antonio Andrea Gargiulo, @crisostomi.bsky.social , @mariasofiab.bsky.social , @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà
Paper: arxiv.org/abs/2412.00081
Code: github.com/AntoAndGar/t...
#machinelearning
Paper: arxiv.org/abs/2412.00081
Code: github.com/AntoAndGar/t...
#machinelearning
January 15, 2025 at 9:36 AM
"Task Singular Vectors: Reducing Task Interference in Model Merging" by Antonio Andrea Gargiulo, @crisostomi.bsky.social , @mariasofiab.bsky.social , @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà
Paper: arxiv.org/abs/2412.00081
Code: github.com/AntoAndGar/t...
#machinelearning
Paper: arxiv.org/abs/2412.00081
Code: github.com/AntoAndGar/t...
#machinelearning
Reposted by Maria Sofia Bucarelli
Don’t miss out on these insights and more — check out the paper!
📄 Preprint → arxiv.org/abs/2412.00081
💻 Code → github.com/AntoAndGar/t...
Joint work w/ Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
(6/6)
📄 Preprint → arxiv.org/abs/2412.00081
💻 Code → github.com/AntoAndGar/t...
Joint work w/ Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
(6/6)
Task Singular Vectors: Reducing Task Interference in Model Merging
Task Arithmetic has emerged as a simple yet effective method to merge models without additional training. However, by treating entire networks as flat parameter vectors, it overlooks key structural in...
arxiv.org
January 8, 2025 at 7:00 PM
Don’t miss out on these insights and more — check out the paper!
📄 Preprint → arxiv.org/abs/2412.00081
💻 Code → github.com/AntoAndGar/t...
Joint work w/ Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
(6/6)
📄 Preprint → arxiv.org/abs/2412.00081
💻 Code → github.com/AntoAndGar/t...
Joint work w/ Antonio A. Gargiulo, @mariasofiab.bsky.social, @sscardapane.bsky.social, Fabrizio Silvestri, Emanuele Rodolà.
(6/6)
Reposted by Maria Sofia Bucarelli
📢Prepend “Singular” to “Task Vectors” and get +15% average accuracy for free!
1. Perform a low-rank approximation of layer-wise task vectors.
2. Minimize task interference by orthogonalizing inter-task singular vectors.
🧵(1/6)
1. Perform a low-rank approximation of layer-wise task vectors.
2. Minimize task interference by orthogonalizing inter-task singular vectors.
🧵(1/6)
January 8, 2025 at 7:00 PM
📢Prepend “Singular” to “Task Vectors” and get +15% average accuracy for free!
1. Perform a low-rank approximation of layer-wise task vectors.
2. Minimize task interference by orthogonalizing inter-task singular vectors.
🧵(1/6)
1. Perform a low-rank approximation of layer-wise task vectors.
2. Minimize task interference by orthogonalizing inter-task singular vectors.
🧵(1/6)
🚀Calling ML on Graphs Enthusiasts! I'm co-organizign a Competition on Learning with Noisy Graph Labels at IJCNN 2025.
Submissions are now open sites.google.com/view/learnin...
Submission Deadline: 10/02/2025
We can’t wait to see your creative approaches and groundbreaking ideas.✨
Submissions are now open sites.google.com/view/learnin...
Submission Deadline: 10/02/2025
We can’t wait to see your creative approaches and groundbreaking ideas.✨
Sign in - Google Accounts
sites.google.com
December 23, 2024 at 4:10 PM
🚀Calling ML on Graphs Enthusiasts! I'm co-organizign a Competition on Learning with Noisy Graph Labels at IJCNN 2025.
Submissions are now open sites.google.com/view/learnin...
Submission Deadline: 10/02/2025
We can’t wait to see your creative approaches and groundbreaking ideas.✨
Submissions are now open sites.google.com/view/learnin...
Submission Deadline: 10/02/2025
We can’t wait to see your creative approaches and groundbreaking ideas.✨