❝HN 片单

Compression for AGI - Jack Rae | Stanford MLSys #76

Stanford MLSys Seminars · 2 位 HN 用户推荐 · 2023–2026

  1. Lossless is actually the appropriate test. These models output probabilities and can easily be turned into very effective lossless compressors with the use of arithmetic encoding. For a very interesting discussion see: [link]
    — modeless, 2023
  2. an LLM can be used to losslessly compress a string to a size equal to the number of bits of entropy of next token prediction loss over the string, by encoding the extra bits of entropy with arithmetic encoding. its sota compression for the distribution of string found on the internet an insightful video on the topic: [link]
    — semiinfinitely, 2026

← HN 片单