Research map / Label-efficient and robust learning / OOD and anomaly detection
LAPT: Label-driven Automated Prompt Tuning for OOD Detection with Vision-Language Models
Yabin Zhang, Wenjie Zhu, Chenhang He, Lei Zhang
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 |
|---|---|---|---|
| replace hand-crafted prompts with learned label-driven prompts | NegLabel · ImageNet-1k, 4 OOD datasets average | FPR95: 25.40 -> 23.40 | Improved |
| replace hand-crafted prompts with LAPT on near-OOD | NegLabel · OpenOOD near-OOD | AUROC: 76.92 -> 82.63 | Improved |
| replace hand-crafted prompts with LAPT on far-OOD | NegLabel · OpenOOD far-OOD | AUROC: 93.30 -> 94.26 | Improved |
| learned prompts also raise ID classification accuracy | NegLabel · OpenOOD ImageNet-1k | ID accuracy: 66.82 -> 67.86 | Improved |
| use distribution-aware prompts instead of unified or class-specific prompts (ablation) | LAPT · OpenOOD | OOD detection | Improved |
| use longer learnable prompts (ablation) | LAPT · OpenOOD | OOD detection | Hurt |
| initialize prompts with 'The nice' instead of random (ablation) | LAPT · OpenOOD | OOD detection | Mixed |
| move class label position within the prompt (ablation) | LAPT · OpenOOD | OOD detection | No clear effect |
| use synthetic images instead of retrieved real images for training (ablation) | LAPT · OpenOOD | OOD detection | Improved |
| combine synthetic and real images (ablation) | LAPT · OpenOOD | OOD detection | Improved |
| use a stronger text-to-image generator (ablation) | LAPT · OpenOOD | OOD detection | Improved |
| use larger-scale WebData for retrieval (ablation) | LAPT · OpenOOD | OOD detection | Improved |
Later papers that compared with this method
2 later accepted papers report LAPT as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- A Close Look at Negative Label Guided Out-of-distribution Detection in Pre-trained Vision-Language Models (ICML 2026 · better on 2 of 2 shared results)
- Overcoming Shortcut Problem in VLM for Robust Out-of-Distribution Detection (CVPR 2025 · better on 2 of 2 shared results)
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
- Conjugated Semantic Pool Improves OOD Detection with Pre-trained Vision-Language Models (NeurIPS 2024 · similarity 0.91)
- Adaptive Multi-prompt Contrastive Network for Few-shot Out-of-distribution Detection (ICML 2025 · similarity 0.90)
- Self-Calibrated Tuning of Vision-Language Models for Out-of-Distribution Detection (NeurIPS 2024 · similarity 0.90)
- Enhanced OoD Detection through Cross-Modal Alignment of Multi-Modal Representations (CVPR 2025 · similarity 0.90)
- Local-Prompt: Extensible Local Prompts for Few-Shot Out-of-Distribution Detection (ICLR 2025 · similarity 0.89)
- DeCoOp: Robust Prompt Tuning with Out-of-Distribution Detection (ICML 2024 · similarity 0.89)
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.
