by Henry | Jul 24, 2024 | entropy, interpretability, publications
Representations as Language An Information Theoretic Framework for Interpretability This paper appeared as a talk at the International Meeting of the Cognitive Science Society in 2024. view paper view code Abstract Large scale neural models show impressive performance...
by Henry | May 1, 2023 | entropy, interpretability, publications
Compositionality With Variation Reliably Emerges in Neural Networks Emergent representations in a multi-agent model are rife with the kinds of variation ubiquitous across natural languages. This paper appeared at ICLR 2024. view paper Abstract We re-evaluated how to...
by Henry | Jul 24, 2021 | optimisation, publications
Meta Learning to Compositionally Generalise Introducing domain-general biases via optimisation This paper appeared as a talk at the Meeting of the Association of Compositional Linguistics (ACL) in 2024. view paper Abstract Natural language is compositional; the...
