Can Performant LLMs Be Ethical? Quantifying the Impact of Web Crawling Opt-Outs

• URL: http://arxiv.org/abs/2504.06219v1
• Date: Tue, 08 Apr 2025 17:08:06 GMT
• Title: Can Performant LLMs Be Ethical? Quantifying the Impact of Web Crawling Opt-Outs
• Authors: Dongyang Fan, Vinko Sabolčec, Matin Ansaripour, Ayush Kumar Tarun, Martin Jaggi, Antoine Bosselut, Imanol Schlag,
• Abstract summary: We quantify the performance difference between models trained on datasets that comply with web crawling opt-outs, and those that do not.<n>Our experiments with 1.5B models show that, as of January 2025, compliance with web data opt-outs does not degrade general knowledge acquisition.<n>However, in specialized domains such as biomedical research, excluding major publishers leads to performance declines.
• Score: 42.58914814153536
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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.

Related papers

• The interplay between domain specialization and model size [8.653321928148547]
  We investigate the interplay between domain and model size during continued pretraining under compute-constrained scenarios.<n>Our goal is to identify an optimal training regime for this scenario and detect patterns in this interplay that can be generalized across different model sizes and domains.
  arXiv  Detail & Related papers  (2025-01-03T19:28:53Z)
• Generate to Discriminate: Expert Routing for Continual Learning [59.71853576559306]
  Generate to Discriminate (G2D) is a continual learning method that leverages synthetic data to train a domain-discriminator.<n>We observe that G2D outperforms competitive domain-incremental learning methods on tasks in both vision and language modalities.
  arXiv  Detail & Related papers  (2024-12-22T13:16:28Z)
• TransformLLM: Adapting Large Language Models via LLM-Transformed Reading Comprehension Text [5.523385345486362]
  We have developed language models specifically designed for legal applications. Our innovative approach significantly improves capabilities in legal tasks by using Large Language Models (LLMs) to convert raw training data into reading comprehension text.
  arXiv  Detail & Related papers  (2024-10-28T19:32:18Z)
• CAP: Detecting Unauthorized Data Usage in Generative Models via Prompt Generation [1.6141139250981018]
  Copyright Audit via Prompts generation (CAP) is a framework for automatically testing whether an ML model has been trained with unauthorized data. Specifically, we devise an approach to generate suitable keys inducing the model to reveal copyrighted contents. To prove its effectiveness, we conducted an extensive evaluation campaign on measurements collected in four IoT scenarios.
  arXiv  Detail & Related papers  (2024-10-08T08:49:41Z)
• Data Shapley in One Training Run [88.59484417202454]
  Data Shapley provides a principled framework for attributing data's contribution within machine learning contexts. Existing approaches require re-training models on different data subsets, which is computationally intensive. This paper introduces In-Run Data Shapley, which addresses these limitations by offering scalable data attribution for a target model of interest.
  arXiv  Detail & Related papers  (2024-06-16T17:09:24Z)
• FairDeDup: Detecting and Mitigating Vision-Language Fairness Disparities in Semantic Dataset Deduplication [28.495688931328882]
  We introduce an easy-to-implement modification to the recent SemDeDup algorithm that can reduce the negative effects that we observe. We find our proposed FairDeDup algorithm consistently leads to improved fairness metrics over SemDeDup on the FairFace and FACET datasets.
  arXiv  Detail & Related papers  (2024-04-24T18:28:17Z)
• The Frontier of Data Erasure: Machine Unlearning for Large Language Models [56.26002631481726]
  Large Language Models (LLMs) are foundational to AI advancements. LLMs pose risks by potentially memorizing and disseminating sensitive, biased, or copyrighted information. Machine unlearning emerges as a cutting-edge solution to mitigate these concerns.
  arXiv  Detail & Related papers  (2024-03-23T09:26:15Z)
• Adapting Large Language Models for Content Moderation: Pitfalls in Data Engineering and Supervised Fine-tuning [79.53130089003986]
  Large Language Models (LLMs) have become a feasible solution for handling tasks in various domains. In this paper, we introduce how to fine-tune a LLM model that can be privately deployed for content moderation.
  arXiv  Detail & Related papers  (2023-10-05T09:09:44Z)
• SILO Language Models: Isolating Legal Risk In a Nonparametric Datastore [159.21914121143885]
  We present SILO, a new language model that manages this risk-performance tradeoff during inference. SILO is built by (1) training a parametric LM on Open License Corpus (OLC), a new corpus we curate with 228B tokens of public domain and permissively licensed text. Access to the datastore greatly improves out of domain performance, closing 90% of the performance gap with an LM trained on the Pile.
  arXiv  Detail & Related papers  (2023-08-08T17:58:15Z)
• Forecasting Workload in Cloud Computing: Towards Uncertainty-Aware Predictions and Transfer Learning [1.5749416770494704]
  We show that modelling the uncertainty of predictions has a positive impact on performance. We investigate whether our models benefit transfer learning capabilities across different domains.
  arXiv  Detail & Related papers  (2023-02-24T14:51:30Z)
• Data-Centric Machine Learning in the Legal Domain [0.2624902795082451]
  This paper explores how changes in a data set influence the measured performance of a model. Using three publicly available data sets from the legal domain, we investigate how changes to their size, the train/test splits, and the human labelling accuracy impact the performance. The observed effects are surprisingly pronounced, especially when the per-class performance is considered.
  arXiv  Detail & Related papers  (2022-01-17T23:05:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.