Pith. sign in

REVIEW 1 cited by

Unmasking and Improving Data Credibility: A Study with Datasets for Training Harmless Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2311.11202 v2 pith:YXMEXGXS submitted 2023-11-19 cs.LG cs.AIcs.CLcs.CY

classification cs.LGcs.AIcs.CLcs.CY
keywords datasetscredibilitydatalanguagecleaningerrorsharmlesslabel
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Language models have shown promise in various tasks but can be affected by undesired data during training, fine-tuning, or alignment. For example, if some unsafe conversations are wrongly annotated as safe ones, the model fine-tuned on these samples may be harmful. Therefore, the correctness of annotations, i.e., the credibility of the dataset, is important. This study focuses on the credibility of real-world datasets, including the popular benchmarks Jigsaw Civil Comments, Anthropic Harmless & Red Team, PKU BeaverTails & SafeRLHF, that can be used for training a harmless language model. Given the cost and difficulty of cleaning these datasets by humans, we introduce a systematic framework for evaluating the credibility of datasets, identifying label errors, and evaluating the influence of noisy labels in the curated language data, specifically focusing on unsafe comments and conversation classification. With the framework, we find and fix an average of 6.16% label errors in 11 datasets constructed from the above benchmarks. The data credibility and downstream learning performance can be remarkably improved by directly fixing label errors, indicating the significance of cleaning existing real-world datasets. We provide an open-source tool, Docta, for data cleaning at https://github.com/Docta-ai/docta.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. When VLMs Meet Image Classification: Test Sets Renovation via Missing Label Identification

    cs.CV 2025-05 conditional novelty 5.0 of 10

    REVEAL ensembles four VLMs and label-noise detectors to detect and correct noisy and missing labels in six image classification test sets, reporting high agreement with human annotations.

Pith tools