Eric Schulz

Reference-Free Rating of LLM Responses via Latent Information

When We Don’t See the Same Picture: Aligning Agents with Divergent Visual Spaces

Exploring System-1 and System-2 Communication for Latent Reasoning in LLMs

A circuit for predicting hierarchical structure in-context in Large Language Models

metabeta - A fast neural model for Bayesian mixed-effects regression

Can vision language models learn intuitive physics from interaction?

Generating Computational Cognitive models using Large Language Models

Concept-Guided Interpretability via Neural Chunking

Building, Reusing, and Generalizing Abstract Representations from Concrete Sequences

metabench - A Sparse Benchmark of Reasoning and Knowledge in Large Language Models

Sparse Autoencoders Reveal Temporal Difference Learning in Large Language Models

Testing the Limits of Fine-Tuning for Improving Visual Cognition in Vision Language Models

Next state prediction gives rise to entangled, yet compositional representations of objects