Learning Task-Agnostic Representations through Multi-Teacher Distillation

Casting complex inputs into tractable representations is a critical step across various fields. Diverse embedding models emerge from differences in architectures, loss functions, input modalities and datasets, each capturing unique aspects of the input. Multi-teacher distillation leverages this diversity to enrich representations but often remains tailored to specific tasks. We introduce a task-agnostic framework based on a ``majority vote" objective function. We demonstrate that this function is bounded by the mutual information between the student and the teachers' embeddings, leading to a task-agnostic distillation loss that eliminates dependence on task-specific labels or prior knowledge. Comprehensive evaluations across text, vision models, and molecular modeling show that our method effectively leverages teacher diversity, resulting in representations enabling better performance for a wide range of downstream tasks such as classification, clustering, or regression. Additionally, we train and release state-of-the-art embedding models, enhancing downstream performance in various modalities.

Authors

• Philippe Formont
• Maxime Darrin International Laboratory on Learning Systems MILA - Quebec AI Institute McGill University Université Paris-Saclay
• Banafsheh Karimian
• Eric Granger ETS Montreal
• Jackie CK Cheung
• Ismail Ayed
• Mohammadhadi Shateri
• Pablo Piantanida International Laboratory on Learning Systems MILA - Quebec AI Institute Université Paris-Saclay CNRS

Keywords

No keywords are indexed for this paper.