{
  "id": 7924105,
  "title": "Objective vs. Search: Decomposing What Makes a Good Tokeniser",
  "url": "https://urgent.news/2026/09/16/objective-vs-search-decomposing-what-makes-a-good-tokeniser",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-16T17:59:45.000Z",
  "source": {
    "name": "arXiv cs.AI",
    "slug": "arxiv-cs-ai",
    "url": "https://arxiv.org/abs/2609.19145v1"
  },
  "original_language": "en",
  "account": null,
  "summary": "Two dominant tokenisation algorithms are used by modern language models: byte-pair encoding (BPE) and UnigramLM. These differ along two orthogonal axes: their optimisation objective (compression vs. log-likelihood) and their search procedure (bottom-up merging vs. top-down pruning). Existing comparisons confound these axes, making it unclear whether their observed differences stem from what is…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}