Twitxy Saved

Denny Zhou

Verified account This account is verified. Learn more

@denny_zhou

@Meta Superintelligence. Prev: Founded & led the Reasoning Team at @Google Brain / Google DeepMind (2021-2026), co-founded COLM
Joined August 2013
597 Following 40K Followers 952 Posts

Latest posts 5

Pinned post

Denny Zhou Verified account @denny_zhou
Slides for my lecture “LLM Reasoning” at Stanford CS 25: dennyzhou.github.io/LLM-Reasoning-…

Key points:
1. Reasoning in LLMs simply means generating a sequence of intermediate tokens before producing the final answer. Whether this resembles human reasoning is irrelevant. The crucial insight is that transformer models can become nearly arbitrarily powerful by generating many intermediate tokens, without the need of scaling the model size (arxiv.org/abs/2402.12875).

2. Pretrained models, even without any fine-tuning, are capable of reasoning. The challenge is that reasoning-based outputs often don’t appear at the top of the output distribution, so standard greedy decoding fails to surface them (arxiv.org/abs/2402.10200)

3. Prompting techniques (e.g., chain-of-thought prompting or "let’s think step by step") and supervised finetuning were commonly used to elicit reasoning. Now, RL finetuning has emerged as the most powerful method. This trick was independently discovered by several labs. At Google, credit goes to Jonathan Lai on my team. Based on our theory ( see point 1), scaling RL should focus on generating long responses rather than something else.

4. LLM reasoning can be hugely improved by generating multiple responses and then aggregating them, rather than relying on a single response (arxiv.org/abs/2203.11171).
62 replies 483 reposts 3.1K likes 470K views View on X
Denny Zhou Verified account @denny_zhou
We still understand surprisingly little about why LLMs work so well, particularly why they can reason so effectively.
Andrew Curran Verified account @AndrewCurran_ Oct 10
This is a slide from Terence Tao's 'Math 2.0' public lecture yesterday evening at Caltech. I am highly confident that, if the cure actually worked, almost any Stage 4 cancer patient would tell you that human comprehension of its mechanism is completely irrelevant to them.
View quoted post on X
6 replies 0 reposts 27 likes 3.7K views View on X