Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.08759 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918437986500608 |
|---|---|
| author | Huang, Yu-Shin Tian, Chao Narayanan, Krishna |
| author_facet | Huang, Yu-Shin Tian, Chao Narayanan, Krishna |
| contents | This paper considers the problem of multi-bit generative watermarking for large language models under a worst-case false-alarm constraint. Prior work established a lower bound on the achievable miss-detection probability in the finite-token regime and proposed a scheme claimed to achieve this bound. We show, however, that the proposed scheme is in fact suboptimal. We then develop two new encoding-decoding constructions that attain the previously established lower bound, thereby completely characterizing the optimal multi-bit watermarking performance. Our approach formulates the watermark design problem as a linear program and derives the structural conditions under which optimality can be achieved. In addition, we identify the failure mechanism of the previous construction and compare the tradeoffs between the two proposed schemes. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_08759 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Optimal Multi-bit Generative Watermarking Schemes Under Worst-Case False-Alarm Constraints Huang, Yu-Shin Tian, Chao Narayanan, Krishna Information Theory Computation and Language This paper considers the problem of multi-bit generative watermarking for large language models under a worst-case false-alarm constraint. Prior work established a lower bound on the achievable miss-detection probability in the finite-token regime and proposed a scheme claimed to achieve this bound. We show, however, that the proposed scheme is in fact suboptimal. We then develop two new encoding-decoding constructions that attain the previously established lower bound, thereby completely characterizing the optimal multi-bit watermarking performance. Our approach formulates the watermark design problem as a linear program and derives the structural conditions under which optimality can be achieved. In addition, we identify the failure mechanism of the previous construction and compare the tradeoffs between the two proposed schemes. |
| title | Optimal Multi-bit Generative Watermarking Schemes Under Worst-Case False-Alarm Constraints |
| topic | Information Theory Computation and Language |
| url | https://arxiv.org/abs/2604.08759 |