Research map / Large language models / Language, safety and interpretability
Learn-to-Distance: Distance Learning for Detecting LLM-Generated Text
Hongyi Zhou, Jin Zhu, Kai Ye, Ying Yang, Erhan Xu, Chengchun Shi
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 |
|---|---|---|---|
| learn the distance function instead of fixed rewrite distances (RAIDAR) | rewrite-based detector · GPT-3.5 Turbo text, 21 domains | AUC: 0.745 (RAIDAR) -> 0.948 | Improved |
| use a learned rewrite distance instead of ML logits-based ImBD | ImBD · GPT-3.5 Turbo text, 21 domains | AUC: 0.890 -> 0.948 | Improved |
| use L2D versus zero-shot FDGPT per domain | FDGPT · Gemini 1.5 Pro text, per domain | AUC: Code 0.829 -> 0.979; ArtCulture 0.982 -> 0.955 | Mixed |
| use L2D on unseen prompts (rewrite, polish, expand) | ImBD · News, GPT-4o | AUC: 0.943 -> 0.998 | Improved |
| use L2D on Wiki with Claude-3.5 | ImBD · Wiki, Claude-3.5 | AUC: 0.961 -> 0.950 | Hurt |
| add a learned distance (fine-tuned LM) vs fixed pre-trained model distance (FD) (ablation) | L2D · same settings as unseen-prompt experiments | AUC: average improvement of 97.1% | Improved |
| paraphrasing or decoherence attacks on LLM text | L2D · News, Wiki, Story | AUC | No clear effect |
| compare runtime to RAIDAR | L2D · GPT-3.5 Turbo text | runtime: slight increase over RAIDAR | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- RADAR — RADAR: Robust AI-Text Detection via Adversarial Learning (NeurIPS 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Training-free LLM-generated Text Detection by Mining Token Probability Sequences (ICLR 2025 · similarity 0.91)
- DALD: Improving Logits-based Detector without Logits from Black-box LLMs (NeurIPS 2024 · similarity 0.88)
- AdaDetectGPT: Adaptive Detection of LLM-Generated Text with Statistical Guarantees (NeurIPS 2025 · similarity 0.88)
- DetectRL: Benchmarking LLM-Generated Text Detection in Real-World Scenarios (NeurIPS 2024 · similarity 0.88)
- Detecting Pretraining Data from Large Language Models (ICLR 2024 · similarity 0.87)
- DNA-GPT: Divergent N-Gram Analysis for Training-Free Detection of GPT-Generated Text (ICLR 2024 · similarity 0.87)
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.
