{
  "id": 1850904,
  "title": "Grading Needs a Rubric, Not Intelligence",
  "url": "https://urgent.news/2026/08/18/grading-needs-a-rubric-not-intelligence",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-18T16:00:05.000Z",
  "source": {
    "name": "arXiv cs.AI",
    "slug": "arxiv-cs-ai",
    "url": "https://arxiv.org/abs/2608.17938v1"
  },
  "original_language": "en",
  "account": null,
  "summary": "Small language models can grade open-ended examination answers as reliably as substantially more expensive models when they grade against an explicit rubric. We test this claim as the design principle behind any-to-bench: a frontier model reads source documents once, at ingestion, to extract each question and its rubric; lower-cost models then perform all repeated grading work. We evaluate six…",
  "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."
}