Undetectable Watermarks for Language Models · HackerPapers discussions

Undetectable Watermarks for Language Models — mentioned by Akranazon

Sun, 01 Mar 2026 08:26:57 GMT

Detecting LLM-generated text is basically solved by modern watermarking techniques (https://arxiv.org/abs/2306.09194). However, the main trouble with watermark-based approaches is that you have to get every LLM provider to adopt it. A student trying to cheat could always opt for some open-weight Chinese model, if the word spreads that the major providers are compromised.

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

Undetectable Watermarks for Language Models — HN submission

Fri, 30 Jun 2023 23:03:18 GMT

Undetectable Watermarks for LLMs

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

Undetectable Watermarks for Language Models — HN submission

Sat, 17 Jun 2023 14:33:29 GMT

Undetectable Watermarks for Language Models

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

Undetectable Watermarks for Language Models — HN submission

Sat, 17 Jun 2023 01:19:12 GMT

Undetectable Watermarks for LLMs

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

Undetectable Watermarks for Language Models — HN submission

Fri, 16 Jun 2023 01:40:47 GMT

Undetectable Watermarks for Language Models

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