The Illusion of Thinking: What the Apple AI Paper Says About LLM Reasoning
Deep Papers
00:00
Comparative Analysis of LLMs and LRMs in Puzzle Complexity
This chapter analyzes the performance of Large Language Models and Logic Reasoning Models on complex puzzles like the Tower of Hanoi and Blocksworld. It highlights how LLMs excel at low complexity while LRMs struggle with higher complexity tasks, with a focus on the 'pass at K' metric for evaluating success.
Transcript
Play full episode
