←back to thread

Karpathy on DeepSeek-OCR paper: Are pixels better inputs to LLMs than text?

(twitter.com)
237 points JnBrymn | 3 comments | 21 Oct 25 17:43 UTC | HN request time: 0.746s | source
Show context
varispeed ◴[22 Oct 25 22:46 UTC] No.45676118[source]▶
Text is linear, whereas image is parallel. I mean when people often read they don't scan text from left to right (or different direction, depending on language), but rather read the text all at once or non-linearly. Like first lock on keywords and then read adjacent words to get meaning, often even skipping some filler sentences unconsciously.

Sequential reading of text is very inefficient.

replies(4): >>45676232 #>>45676919 #>>45677443 #>>45677649 #
1. jb1991 ◴[23 Oct 25 02:47 UTC] No.45677649[source]▶
I think you’re making a lot of assumptions about how people read.
replies(1): >>45678093 #
2. com2kid ◴[23 Oct 25 04:11 UTC] No.45678093[source]▶
He isn't, plenty of studies have been done on the topic. Eyes dart around a lot when reading.
replies(1): >>45678260 #
3. jb1991 ◴[23 Oct 25 04:52 UTC] No.45678260[source]▶
People do skip words or scan for key phrases, but reading still happens in sequence. The brain depends on word order and syntax to make sense of text, so you cannot truly read it all at once. Skimming just means you sample parts of a linear structure, not that reading itself is non-linear. Eye-tracking studies confirm this sequential processing (check out the Rayner study in Psychological Bulletin if you are interested).