Himabindu Lakkaraju

New methods improve how machine learning handles noisy data.

An exploration of how CoT prompting influences language model behavior and performance.

Introducing Verifiability Tuning for clearer and trustworthy AI explanations.

Addressing privacy risks while providing valuable insights from machine learning models.

A new method enhances the safety of language models against harmful prompts.

Examining how robust machine learning models impact explanation effectiveness.

Examining the challenges of self-explanations in large language models.

SpLiCE helps clarify the dense data from CLIP for better understanding.

Assessing how data poisoning affects policy evaluation methods.

Understanding AI decision-making is crucial for trust and ethical use.

Examining the effectiveness of reasoning in large language models.

Regulations guide the safe and fair use of AI technologies across various sectors.

Research reveals similarities in image models' internal representations.

Examining privacy risks in model explanations and strategies to enhance security.

A method to understand user preferences for modifying machine learning outcomes.

A straightforward look at large language models and their workings.

Exploring how fine-tuning affects reasoning in language models.