Research map / Large language models / Language, safety and interpretability
Protecting Intellectual Property of Language Generation APIs with Lexical Watermark
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add lexical watermarking through synonym or spelling replacements | NLG API translation · Translation | BERTScore: around 94.3 -> around 94.2 | No clear effect |
| Add lexical watermarking | NLG API outputs copied by an imitation attacker · Imitation attack using watermarked data | watermark detectability | Helped |
| Bias lexical word selection in the API outputs | Imitation attack model · Imitation attack | watermark detection | Helped |
Closest papers by meaning
- Can Watermarking Large Language Models Prevent Copyrighted Text Generation and Hide Training Data? (AAAI 2025 · similarity 0.87)
- A Semantic Invariant Robust Watermark for Large Language Models (ICLR 2024 · similarity 0.85)
- Unbiased Watermark for Large Language Models (ICLR 2024 · similarity 0.85)
- Adaptive Text Watermark for Large Language Models (ICML 2024 · similarity 0.85)
- On the Learnability of Watermarks for Language Models (ICLR 2024 · similarity 0.85)
- Watermarking Conditional Text Generation for AI Detection: Unveiling Challenges and a Semantic-Aware Watermark Remedy (AAAI 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.