21 citations · 31 across the 5 of their papers we have counts for
5 papers · 1 filter
Next state prediction gives rise to entangled, yet compositional representations of objects
Tankred Saanum, Luca M. Schulze Buschoff, Peter Dayan +1
Compositional representations are thought to enable humans to generalize across combinatorially vast state spaces. Models with learnable object slots, which encode information abou…
Sparse Autoencoders Reveal Temporal Difference Learning in Large Language Models
Can Demircan, Tankred Saanum, Akshay K. Jagadish +2
In-context learning, the ability to adapt based on a few examples in the input prompt, is a ubiquitous feature of large language models (LLMs). However, as LLMs' in-context learnin…
In-context learning agents are asymmetric belief updaters
Johannes A. Schubert, Akshay K. Jagadish, Marcel Binz +1
We study the in-context learning dynamics of large language models (LLMs) using three instrumental learning tasks adapted from cognitive psychology. We find that LLMs update their…
The Acquisition of Physical Knowledge in Generative Neural Networks
Luca M. Schulze Buschoff, Eric Schulz, Marcel Binz
As children grow older, they develop an intuitive understanding of the physical processes around them. Their physical understanding develops in stages, moving along developmental t…
Reinforcement Learning with Simple Sequence Priors
Tankred Saanum, Noémi Éltető, Peter Dayan +2
Everything else being equal, simpler models should be preferred over more complex ones. In reinforcement learning (RL), simplicity is typically quantified on an action-by-action ba…