Can Performant LLMs Be Ethical? Quantifying the Impact of Web Crawling Opt-Outs
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913975335124992 |
|---|---|
| author | Fan, Dongyang Sabolčec, Vinko Ansaripour, Matin Tarun, Ayush Kumar Jaggi, Martin Bosselut, Antoine Schlag, Imanol |
| author_facet | Fan, Dongyang Sabolčec, Vinko Ansaripour, Matin Tarun, Ayush Kumar Jaggi, Martin Bosselut, Antoine Schlag, Imanol |
| contents | The increasing adoption of web crawling opt-outs by copyright holders of online content raises critical questions about the impact of data compliance on large language model (LLM) performance. However, little is known about how these restrictions (and the resultant filtering of pretraining datasets) affect the capabilities of models trained using these corpora. In this work, we conceptualize this effect as the $\textit{data compliance gap}$ (DCG), which quantifies the performance difference between models trained on datasets that comply with web crawling opt-outs, and those that do not. We measure the data compliance gap in two settings: pretraining models from scratch and continual pretraining from existing compliant models (simulating a setting where copyrighted data could be integrated later in pretraining). Our experiments with 1.5B models show that, as of January 2025, compliance with web data opt-outs does not degrade general knowledge acquisition (close to 0\% DCG). However, in specialized domains such as biomedical research, excluding major publishers leads to performance declines. These findings suggest that while general-purpose LLMs can be trained to perform equally well using fully open data, performance in specialized domains may benefit from access to high-quality copyrighted sources later in training. Our study provides empirical insights into the long-debated trade-off between data compliance and downstream model performance, informing future discussions on AI training practices and policy decisions. Our website is available at https://data-compliance.github.io/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_06219 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Can Performant LLMs Be Ethical? Quantifying the Impact of Web Crawling Opt-Outs Fan, Dongyang Sabolčec, Vinko Ansaripour, Matin Tarun, Ayush Kumar Jaggi, Martin Bosselut, Antoine Schlag, Imanol Computation and Language Machine Learning The increasing adoption of web crawling opt-outs by copyright holders of online content raises critical questions about the impact of data compliance on large language model (LLM) performance. However, little is known about how these restrictions (and the resultant filtering of pretraining datasets) affect the capabilities of models trained using these corpora. In this work, we conceptualize this effect as the $\textit{data compliance gap}$ (DCG), which quantifies the performance difference between models trained on datasets that comply with web crawling opt-outs, and those that do not. We measure the data compliance gap in two settings: pretraining models from scratch and continual pretraining from existing compliant models (simulating a setting where copyrighted data could be integrated later in pretraining). Our experiments with 1.5B models show that, as of January 2025, compliance with web data opt-outs does not degrade general knowledge acquisition (close to 0\% DCG). However, in specialized domains such as biomedical research, excluding major publishers leads to performance declines. These findings suggest that while general-purpose LLMs can be trained to perform equally well using fully open data, performance in specialized domains may benefit from access to high-quality copyrighted sources later in training. Our study provides empirical insights into the long-debated trade-off between data compliance and downstream model performance, informing future discussions on AI training practices and policy decisions. Our website is available at https://data-compliance.github.io/. |
| title | Can Performant LLMs Be Ethical? Quantifying the Impact of Web Crawling Opt-Outs |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2504.06219 |