Abhilasha Ravichander

This first study of Web search across four major conversational platforms is presented, combining real-world user interactions with controlled experiments using the same platform's models by their APIs (invitro), finding that Web-search decisions vary substantially across platforms and models.

Moving beyond passive prompt-and-generate workflows to genuine human-AI collaboration requires solving open technical challenges in mutual agency, real-time latent steering, and dynamic evaluation.

You can slash RoPE memory costs by 10x without sacrificing convergence, just by applying it to a sliver (10%) of hidden dimensions.

LLMs exhibit surprisingly strong and predictable biases towards specific information sources, even overriding content relevance and explicit instructions.