Andrej Karpathy(@karpathy)
@kepano I just tried it this morning on the 245-page Mythos pdf and it failed badly and the outputs ...
5.5内容质量
TL;DR · AI 摘要
Andrej Karpathy指出当前SOTA大模型在处理245页PDF时表现糟糕，认为PDF转换应作为LLM的专用技能而非通用程序。
核心要点
- 当前顶尖大模型处理复杂PDF文档仍存在严重问题
- PDF内容提取失败表现为输出混乱、结构错乱
- 建议将PDF解析设计为LLM的专项技能而非通用功能
#大语言模型#PDF处理#AI工程
打开原文Andrej Karpathy on X: "@kepano I just tried it this morning on the 245-page Mythos pdf and it failed badly and the outputs were all mangled. Converting pdfs is really hard, I think it has to probably be a Skill not a program, for a SOTA LLM for it to work properly." / X
Don’t miss what’s happening
I just tried it this morning on the 245-page Mythos pdf and it failed badly and the outputs were all mangled. Converting pdfs is really hard, I think it has to probably be a Skill not a program, for a SOTA LLM for it to work properly.
·
167
61
1.6K
627
Read 167 replies