Tiberiu Musat

According to our database, Tiberiu Musat authored at least 4 papers between 2024 and 2025.

Collaborative distances:

• Dijkstra number of five.
• Erdős number of four.

2025

The Geometry of Grokking: Norm Minimization on the Zero-Loss Manifold.

Tiberiu Musat

CoRR, November, 2025

On the Emergence of Induction Heads for In-Context Learning.

Tiberiu Musat, Tiago Pimentel, Lorenzo Noci, Alessandro Stolfo, Mrinmaya Sachan, Thomas Hofmann

CoRR, November, 2025

Mechanism and Emergence of Stacked Attention Heads in Multi-Layer Transformers.

Tiberiu Musat

Proceedings of the Thirteenth International Conference on Learning Representations, 2025

2024

Clustering and Alignment: Understanding the Training Dynamics in Modular Addition.

Tiberiu Musat

CoRR, 2024