Official data release to reproduce Confident Learning paper results
Use Large Language Models like OpenAI's GPT-3.5 for data annotation and model enhancement. This framework combines human expertise with LLMs, employs Iterative Active Learning for continuous improvement, and integrates CleanLab (Confident Learning) to ensure high-quality datasets and better model performance
Finds mislabelled training data, and measures whether fixing it beats improving the model
Confident Learning + CORN ordinal regression on IndoBERT for handling label noise in Indonesian app review rating prediction. Undergraduate thesis code.
Data-Centric-AI
Empirical investigation of when data quality remediation outperforms architectural upgrades in NLP text classification, across 135 experimental conditions spanning five corpora, three architectures, and four annotation cost tiers.
Can capture-recapture count the benchmark label errors nobody found? Two pre-registered experiments on MMLU-Redux with LLM judges and confident learning (negative result).
To associate your repository with the confident-learning topic, visit your repo's landing page and select "manage topics."