#llm-scaling on Hacker News

2 stories tagged llm-scaling, newest first.

Screenshot of arrowtsx.dev: GPT-5.5 hallucinates 3x more than MIT-licensed GLM-5.2

GPT-5.5 hallucinates 3x more than MIT-licensed GLM-5.2 (arrowtsx.dev)

474 points by oshrimpton 4 months ago | 235 comments

Blog post arguing that bigger AI models are not better, comparing benchmark and hallucination rates of GLM-5.2, GPT-5.5, Opus 4.8, DeepSeek V4 and others.

llm-scaling hallucination glm-5-2 gpt-5-5

Screenshot of jax-ml.github.io: How to Think About GPUs

How to Think About GPUs (jax-ml.github.io)

361 points by alphabetting 1 year ago | 107 comments

Part 12 of Google's "How To Scale Your Model" book, a deep dive into how NVIDIA GPUs work and are networked, and what that means for LLMs compared to TPUs.

gpus nvidia tpus llm-scaling

#llm-scaling on Hacker News

GPT-5.5 hallucinates 3x more than MIT-licensed GLM-5.2(arrowtsx.dev)

How to Think About GPUs(jax-ml.github.io)

GPT-5.5 hallucinates 3x more than MIT-licensed GLM-5.2 (arrowtsx.dev)

How to Think About GPUs (jax-ml.github.io)