Akshay K. Jagadish
Doctoral Researcher, Helmholtz Munich
October 2024 Preprint out on "Sparse Autoencoders Reveal Temporal Difference Learning in Large Language Models" and "Centaur: a foundation model of human cognition"
September 2024 Commentaries and responses for our BBS paper are now available
July 2024 I presented my work on "Zero-shot Compositional Reinforcement Learning" in the Compositionality in minds, brains and machines workshop at CogSci 2024 (Rotterdam)
July 2024 We are organizing a workshop on In-context learning in natural and artificial intelligence at CogSci 2024 (Rotterdam)
July 2024 I will be at ICML in Vienna to present two of our works: "Ecologically rational meta-learned inference explains human category learning" and "In-context learning agents are asymmetric belief updaters"
June 2024 I presented my work on "Ecologically Rational Meta-learned Inference" at the Indian Institue of Science, Bangalore
May 2024 Preprint out on “Chat-GPT on the Couch”: Assessing and Alleviating State Anxiety in Large Language Models