Is Thinking an Illusion?

Research Overview

Large Language Models (LLMs) show remarkable capabilities, but whether they genuinely reason or merely recite memorized patterns remains an open question. This project introduces a novel framework to distinguish these behaviors using competitive programming problems as a verifiable testbed.

Status

This was initially an ANLP (Advanced NLP) course project at IIIT Hyderabad. The research is currently being expanded for potential publication.