Research map / Large language models / Language, safety and interpretability
A Content-Preserving Secure Linguistic Steganography
Lingyun Xiang, Chengfu Ou, Xu He, Zhongliang Yang, Yuling Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| use CLstega instead of baselines | BERT-based steganography · BiLSTM-Dense, SeSy, HiDuNet | steganalysis accuracy: Discop 0.5085 -> CLstega 0.4955 (BiLSTM-Dense) | Improved |
| CLstega vs Discop on SeSy | CLstega · SeSy | steganalysis accuracy: 0.5032 -> 0.5038 | No clear effect |
| use CLstega instead of baselines | CLstega · CC-100 | perplexity: 82.55 -> 70.16 | Improved |
| CLstega vs generation-based methods on capacity | CLstega · CC-100 | embedding rate: 5.5256 -> 0.4204 | Hurt |
| CLstega vs modification-based methods on capacity | CLstega · CC-100 | embedding rate: 0.2542 -> 0.4204 | Improved |
| use single-position augmented masking (SPAM) instead of full-position masking (ablation) | CLstega · extraction success rate | ESR convergence | Improved |
| use SPAM instead of FPM (ablation) | CLstega · extraction time, k=all NFW | extraction time: 0.0302 -> 0.1543 sec | Hurt |
| increase embedding positions k (ablation) | CLstega · extraction success rate | ESR | Hurt |
| embed in all words instead of non-functional words only (ablation) | CLstega · k=all | embedding rate: 0.4204 -> 0.9538 | Improved |
Closest papers by meaning
- Promising Multi-Granularity Linguistic Steganography by Jointing Syntactic and Lexical Manipulations (AAAI 2025 · similarity 0.90)
- STEAD: Robust Provably Secure Linguistic Steganography with Diffusion Language Model (NeurIPS 2025 · similarity 0.90)
- STLC-KG:A Social Text Steganalysis Method Combining Large-Scale Language Models and Common-Sense Knowledge Graphs (AAAI 2025 · similarity 0.84)
- A Resilient and Accessible Distribution-Preserving Watermark for Large Language Models (ICML 2024 · similarity 0.84)
- SSHR: More Secure Generative Steganography with High-Quality Revealed Secret Images (ICML 2025 · similarity 0.82)
- Theoretically Grounded Framework for LLM Watermarking: A Distribution-Adaptive Approach (NeurIPS 2025 · similarity 0.82)
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.
