Gary Wang Recommends Video on LLM Thinking and Depth

Gary Wang, who works on speech synthesis, ASR, RL, and NLP for Gemini at Google, posted that a video offers useful insight into how LLMs think both with and without chain-of-thought monitoring. He linked the recommendation to ongoing talks about looping architectures and CoT monitoring, noting the video's value for seeing the relationship between those elements and model depth. Wang's background includes prior work as a chemical engineer on fuel cells, batteries, and oil and gas. The post appears on his account focused on those technical areas.