Towards Data Science

The essence of using sets to boost performance is encapsulated by the principle of optimization by reduction. Read more from Afjal Chowdhury now.

Choosing between one- and two-tailed hypotheses affects every stage of A/B testing. Learn why the hypothesis direction matters and explore the pros and cons of each approach in Allon Korem's article.

Lorenzo Mezzini illustrates how to use Fourier Transform to detect seasonal components.

Matt Chapman introduces SetFit, a fine-tuning framework that can help you build highly performant NLP classifiers with as few as 8 labelled samples per class.

Adam Kovacs dives into LettuceDetect, a lightweight hallucination detector for #RAG pipelines. This encoder-based model is built on #ModernBERT, with ready-to-use #Python packages and pretrained models.

Large language model hallucinations are inherent consequences of the training pipeline, particularly arising from the supervised fine-tuning stage. Read more from Prashal Ruchiranga.

Wonder why #LLMs sometimes struggle with syntax? Learn how to use Sparse Autoencoders to uncover hidden circuits for subject-verb agreement Shuyang Xiang's article.

Video processing for deep learning is often slow, causing latency and #GPU bottlenecks. David Redó Nieto optimizes performance using PyTorch, FFmpeg, and NVIDIA acceleration — making object detection, classification, and segmentation much faster.

Niklas Lang provides a comprehensive overview of Hadoop and its components.

Learn about the risks of data reconstruction attacks and the importance of robust privacy-preserving techniques in medical AI in Jonathan Yahav’s article.