← Back to context Comment by pcwelder 4 months ago I ϲаn guаrаntее thаt thе ОСR ϲаn't rеаd thіs sеntеnсе ϲоrrесtlу. 8 comments pcwelder Reply syntaxing 4 months ago What’s correct though? Even as a human, I read that “correctly”. Using weird representations of C doesn’t change the word? LudwigNagasena 4 months ago I would even say that OCR can rеаd the sеntеnсе ϲоrrесtlу, while a tokenizer can't. kgeist 4 months ago Qwen3 8b perfectly understood it after 14 seconds of thinking. metalliqaz 4 months ago Yeah OCR would be much more likely to read that sentence the way a human would. bitdivision 4 months ago A lot of Cyrillic characters: https://apps.timwhitlock.info/unicode/inspect?s=I+%CF%B2%D0%... geysersam 4 months ago Really? How so? moduspol 4 months ago Looks like he’s using atypical “c” characters.
syntaxing 4 months ago What’s correct though? Even as a human, I read that “correctly”. Using weird representations of C doesn’t change the word? LudwigNagasena 4 months ago I would even say that OCR can rеаd the sеntеnсе ϲоrrесtlу, while a tokenizer can't. kgeist 4 months ago Qwen3 8b perfectly understood it after 14 seconds of thinking. metalliqaz 4 months ago Yeah OCR would be much more likely to read that sentence the way a human would.
LudwigNagasena 4 months ago I would even say that OCR can rеаd the sеntеnсе ϲоrrесtlу, while a tokenizer can't. kgeist 4 months ago Qwen3 8b perfectly understood it after 14 seconds of thinking.
metalliqaz 4 months ago Yeah OCR would be much more likely to read that sentence the way a human would.
bitdivision 4 months ago A lot of Cyrillic characters: https://apps.timwhitlock.info/unicode/inspect?s=I+%CF%B2%D0%...
geysersam 4 months ago Really? How so? moduspol 4 months ago Looks like he’s using atypical “c” characters.
What’s correct though? Even as a human, I read that “correctly”. Using weird representations of C doesn’t change the word?
I would even say that OCR can rеаd the sеntеnсе ϲоrrесtlу, while a tokenizer can't.
Qwen3 8b perfectly understood it after 14 seconds of thinking.
Yeah OCR would be much more likely to read that sentence the way a human would.
A lot of Cyrillic characters: https://apps.timwhitlock.info/unicode/inspect?s=I+%CF%B2%D0%...
Really? How so?
Looks like he’s using atypical “c” characters.