I am a PhD candidate at the Swiss AI Lab IDSIA, University of Lugano, where I work under the supervision of Prof. Jürgen Schmidhuber. My research centers on artificial curiosity, representation learning, reinforcement learning, and information theory, with the aim of uncovering criteria for determining the ‘interestingness’ of objects and developing effective representations for them. In addition to my academic pursuits, I am a classical pianist and composer.
Recent News
Paper accepted at IEEE TPAMI
Our work “On the Distillation of Stories for Transferring Narrative Arcs in Collections of Independent Media” has been accepted as a full paper at IEEE TPAMI...
ICML 2024 Oral
Presented our oral “Learning Useful Representations of Recurrent Neural Network Weight Matrices” at ICML 2024 in Vienna. Presentation Paper Poster
Research Projects
Multiple Token Divergence—Measuring and Steering In-Context Computation Density
The divergence between the prediction distributions of the full model and a computational shortcut allows us to measure and steer the computational effort of...
Measuring In-Context Computation Complexity via Hidden State Prediction
Can we measure when to what degree a sequence model or LLM is doing something interesting? We propose hidden state unpredictability as a measure, and show th...
