Investigating how environmental constraints shape research methodologies
Exploring reward mechanisms in large language models
Understanding how neural networks develop internal representations
Mapping contextual information to neural network parameters
Analyzing LLM performance with selective domain removal
High-dimensional compositional generalization research
Examining the relationship between learned representations and training objectives
Teaching models to express uncertainty appropriately
Reinforcement learning with synthetic environments
Evaluating research capabilities of language models
Large-scale synthetic data generation and evaluation
Connecting reward systems to exploratory behavior
These are publicly shared research projects.
Create an account to share your own research.