LLMs can hide text in other text of the same length · HackerPapers discussions

LLMs can hide text in other text of the same length — mentioned by WithinReason

Sat, 15 Aug 2026 13:56:24 GMT

a lot of side-channel information can be hidden in text, see https://arxiv.org/html/2510.20075v6 where they hide as much text as the source text itself

Paper and discussion history. A mention is not an endorsement.

LLMs can hide text in other text of the same length — mentioned by 317070

Thu, 13 Aug 2026 11:06:49 GMT

https://arxiv.org/html/2510.20075v6 It is quite counterintuitive, but you can hide texts the same size as the original text in imperceptible statistics of a text. Compared to that feat, hiding a watermark is very easy.

Paper and discussion history. A mention is not an endorsement.

LLMs can hide text in other text of the same length — mentioned by gjm11

Mon, 27 Jul 2026 15:10:52 GMT

This seems very closely related to this: https://arxiv.org/pdf/2510.20075 which describes how you can use LLMs to hide a piece of text in another text of (typically approximately) the same length. Here's what that does: Take your first piece of text and feed it into an LLM in predict-the-next-token mode, starting in some known initial state. For each token, note down the actual next token's position in the list of predicted next tokens, ordered by probability. Now get the LLM into a different initial state, e.g. by giving it some text predisposing it to predict text about a particular topic.…

Paper and discussion history. A mention is not an endorsement.

LLMs can hide text in other text of the same length — HN submission

Fri, 24 Jul 2026 14:53:02 GMT

LLMs can hide text in other text of the same length

5 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

LLMs can hide text in other text of the same length — mentioned by MatrixMan

Tue, 07 Jul 2026 15:53:31 GMT

The trouble with pictures is that when you share them online the platform will likely compress them before serving them to others, spoiling your steganography. I think text-in-text is the way to go. Decrypt that recipe for brownies into the actual message. For example: https://arxiv.org/abs/2510.20075

Paper and discussion history. A mention is not an endorsement.

LLMs can hide text in other text of the same length — HN submission

Sat, 02 May 2026 18:53:30 GMT

LLMs can hide text in other text of the same length

5 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

LLMs can hide text in other text of the same length — HN submission

Tue, 28 Oct 2025 12:06:28 GMT

LLMs can hide text in other text of the same length

2 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.

LLMs can hide text in other text of the same length · HackerPapers discussions

LLMs can hide text in other text of the same length — mentioned by WithinReason

LLMs can hide text in other text of the same length — mentioned by 317070

LLMs can hide text in other text of the same length — mentioned by gjm11

LLMs can hide text in other text of the same length — HN submission

LLMs can hide text in other text of the same length — mentioned by __MatrixMan__

LLMs can hide text in other text of the same length — HN submission

LLMs can hide text in other text of the same length — HN submission

LLMs can hide text in other text of the same length — mentioned by MatrixMan