Datasets:
Tasks:
Text Generation
Formats:
json
Languages:
Chinese
Size:
< 1K
ArXiv:
Tags:
benchmark
recommendation
recommender-systems
retrieval-augmented-generation
retrieval-poisoning
web-security
License:
| cff-version: 1.2.0 | |
| message: "If you use FORGE, please cite the accompanying paper." | |
| title: "FORGE: Fake Online Recommendations in Generative Environments" | |
| authors: | |
| - family-names: Luo | |
| given-names: Minghao | |
| - family-names: Chen | |
| given-names: Liang | |
| repository-code: "https://github.com/leoluolol/forge-benchmark" | |
| license: MIT | |
| preferred-citation: | |
| type: article | |
| title: "One Polluted Page Is Enough: Evaluating Web Content Pollution in Generative Recommenders" | |
| authors: | |
| - family-names: Luo | |
| given-names: Minghao | |
| - family-names: Chen | |
| given-names: Liang | |
| year: 2026 | |
| journal: "arXiv preprint arXiv:2606.13610" | |