Tag
An 'Ask HN' post where users share their current tech projects, including a serverless budgeting tool, an AI-first architecture experiment, text compression research, a tiny neural network, and a voxel game engine.
This paper introduces Diffusion Language Models (DLMs) as a new inference paradigm for lossless text compression, aiming to overcome the throughput bottlenecks of autoregressive LLM-based compressors while achieving state-of-the-art compression ratios.
This paper introduces JOLT, an integer programming approach to optimize subword tokenization for greedy left-to-right longest-match decoding (WordPiece). JOLT achieves near-optimal compression, closing most of the gap between BPE and the theoretical lower bound, reducing token count by up to 0.78% over BPE.
This paper systematically benchmarks various deletion strategies (e.g., frequency-guided, entropy-based) for lossy text compression where LLMs reconstruct the original text, showing that simple methods like word-frequency deletion remain competitive across retention rates.