Toxicity of the Commons: Curating Open-Source Pre-Training Data

• URL: http://arxiv.org/abs/2410.22587v2
• Date: Mon, 18 Nov 2024 18:42:44 GMT
• Title: Toxicity of the Commons: Curating Open-Source Pre-Training Data
• Authors: Catherine Arnett, Eliot Jones, Ivan P. Yamshchikov, Pierre-Carl Langlais,
• Abstract summary: We propose a data curation pipeline to reduce harmful outputs by models trained on public domain data. Current state-of-the-art approaches to toxicity filtering are often infeasible or inappropriate for open data models. We create a custom training dataset, ToxicCommons, which is composed of texts which have been classified across five different dimensions.
• Score: 6.137272725645159
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Open-source large language models are becoming increasingly available and popular among researchers and practitioners. While significant progress has been made on open-weight models, open training data is a practice yet to be adopted by the leading open-weight models creators. At the same time, there researchers are working to make language models safer. We propose a data curation pipeline to reduce harmful outputs by models trained on public domain data. There are unique challenges to working with public domain data, as these sources differ from web text in both form and content. Many sources are historical documents and are the result of Optical Character Recognition (OCR). Consequently, current state-of-the-art approaches to toxicity filtering are often infeasible or inappropriate for open data models. In this paper, we introduce a new fully open-source pipeline for open-data toxicity filtering. Our contributions are threefold. We create a custom training dataset, ToxicCommons, which is composed of texts which have been classified across five different dimensions (racial/origin-based, gender/sex-based, religious, ability-based discrimination, and violence). We use this dataset to train a custom classifier, Celadon, that can be used to detect toxic content in open data more efficiently at a larger scale. Finally, we describe the balanced approach to content filtration that optimizes safety filtering with respect to the filtered data available for training.

Related papers

• Approximating Language Model Training Data from Weights [70.08614275061689]
  We formalize the problem of data approximation from model weights and propose several baselines and metrics.<n>We develop a gradient-based approach that selects the highest-matching data from a large public text corpus.<n>Even when none of the true training data is known, our method is able to locate a small subset of public Web documents.
  arXiv  Detail & Related papers  (2025-06-18T15:26:43Z)
• Recycling the Web: A Method to Enhance Pre-training Data Quality and Quantity for Language Models [107.24906866038431]
  We propose REWIRE, REcycling the Web with guIded REwrite, to enrich low-quality documents so that they could become useful for training.<n>We show that mixing high-quality raw texts and our rewritten texts lead to 1.0, 1.3 and 2.5 percentage points improvement respectively across 22 diverse tasks.
  arXiv  Detail & Related papers  (2025-06-05T07:12:12Z)
• OpenCodeReasoning: Advancing Data Distillation for Competitive Coding [61.15402517835137]
  We build a supervised fine-tuning (SFT) dataset to achieve state-of-the-art coding capability results in models of various sizes. Our models use only SFT to achieve 61.8% on LiveCodeBench and 24.6% on CodeContests, surpassing alternatives trained with reinforcement learning.
  arXiv  Detail & Related papers  (2025-04-02T17:50:31Z)
• Scaling Laws for Data Filtering -- Data Curation cannot be Compute Agnostic [99.3682210827572]
  Vision-language models (VLMs) are trained for thousands of GPU hours on carefully curated web datasets. Data curation strategies are typically developed agnostic of the available compute for training. We introduce neural scaling laws that account for the non-homogeneous nature of web data.
  arXiv  Detail & Related papers  (2024-04-10T17:27:54Z)
• Data-Independent Operator: A Training-Free Artifact Representation Extractor for Generalizable Deepfake Detection [105.9932053078449]
  In this work, we show that, on the contrary, the small and training-free filter is sufficient to capture more general artifact representations. Due to its unbias towards both the training and test sources, we define it as Data-Independent Operator (DIO) to achieve appealing improvements on unseen sources. Our detector achieves a remarkable improvement of $13.3%$, establishing a new state-of-the-art performance.
  arXiv  Detail & Related papers  (2024-03-11T15:22:28Z)
• Scalable Extraction of Training Data from (Production) Language Models [93.7746567808049]
  This paper studies extractable memorization: training data that an adversary can efficiently extract by querying a machine learning model without prior knowledge of the training dataset. We show an adversary can extract gigabytes of training data from open-source language models like Pythia or GPT-Neo, semi-open models like LLaMA or Falcon, and closed models like ChatGPT.
  arXiv  Detail & Related papers  (2023-11-28T18:47:03Z)
• Recovering from Privacy-Preserving Masking with Large Language Models [14.828717714653779]
  We use large language models (LLMs) to suggest substitutes of masked tokens. We show that models trained on the obfuscation corpora are able to achieve comparable performance with the ones trained on the original data.
  arXiv  Detail & Related papers  (2023-09-12T16:39:41Z)
• Exploring Data Redundancy in Real-world Image Classification through Data Selection [20.389636181891515]
  Deep learning models often require large amounts of data for training, leading to increased costs. We present two data valuation metrics based on Synaptic Intelligence and gradient norms, respectively, to study redundancy in real-world image data. Online and offline data selection algorithms are then proposed via clustering and grouping based on the examined data values.
  arXiv  Detail & Related papers  (2023-06-25T03:31:05Z)
• A Pretrainer's Guide to Training Data: Measuring the Effects of Data Age, Domain Coverage, Quality, & Toxicity [84.6421260559093]
  This study is the largest set of experiments to validate, quantify, and expose undocumented intuitions about text pretraining. Our findings indicate there does not exist a one-size-fits-all solution to filtering training data.
  arXiv  Detail & Related papers  (2023-05-22T15:57:53Z)
• Bag of Tricks for Training Data Extraction from Language Models [98.40637430115204]
  We investigate and benchmark tricks for improving training data extraction using a publicly available dataset. The experimental results show that several previously overlooked tricks can be crucial to the success of training data extraction.
  arXiv  Detail & Related papers  (2023-02-09T06:46:42Z)
• Perplexed by Quality: A Perplexity-based Method for Adult and Harmful Content Detection in Multilingual Heterogeneous Web Data [0.0]
  We explore different methods for detecting adult and harmful of content in multilingual heterogeneous web data. We train solely with adult and harmful textual data, and then select the documents having a perplexity value above a given threshold. This approach will virtually cluster our documents into two distinct groups, which will greatly facilitate the choice of the threshold for the perplexity.
  arXiv  Detail & Related papers  (2022-12-20T17:14:45Z)
• You Are What You Write: Preserving Privacy in the Era of Large Language Models [2.3431670397288005]
  We present an empirical investigation into the extent of the personal information encoded into pre-trained representations by a range of popular models. We show a positive correlation between the complexity of a model, the amount of data used in pre-training, and data leakage.
  arXiv  Detail & Related papers  (2022-04-20T11:12:53Z)
• WAFFLe: Weight Anonymized Factorization for Federated Learning [88.44939168851721]
  In domains where data are sensitive or private, there is great value in methods that can learn in a distributed manner without the data ever leaving the local devices. We propose Weight Anonymized Factorization for Federated Learning (WAFFLe), an approach that combines the Indian Buffet Process with a shared dictionary of weight factors for neural networks.
  arXiv  Detail & Related papers  (2020-08-13T04:26:31Z)

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.