{
  "id": 4948781,
  "title": "BenchMIRT: What are LLM benchmarks actually measuring?",
  "url": "https://urgent.news/2026/09/01/benchmirt-what-are-llm-benchmarks-actually-measuring",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-01T21:39:07.000Z",
  "source": {
    "name": "Hugging Face",
    "slug": "hugging-face",
    "url": "https://huggingface.co/blog/allenai/benchmirt"
  },
  "original_language": "en",
  "account": "LLM benchmarks often measure more than their intended ability, according to a new method called BenchMIRT. Researchers have developed a way to dissect the individual prompts within a benchmark to reveal what specific skills they are actually assessing. By applying Item Response Theory at both the model and question levels, BenchMIRT can separate multiple capabilities that contribute to a model's performance. This technique has been trained on benchmarking results from 100 LLMs across 16 benchmarks, revealing two dominant dimensions: safety and general reasoning. Some benchmarks, like BBQ which tests social stereotypes, align more with general reasoning than safety. Others, like WMDP which evaluates dangerous knowledge, show a stronger connection to general reasoning. HarmBench, which tests compliance with harmful requests, demonstrates how a single benchmark can incorporate various signals, with both safety and general reasoning dimensions strongly associated. Overall, BenchMIRT highlights that a single benchmark score can combine multiple signals, offering a more nuanced interpretation of a model's abilities.",
  "summary": null,
  "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."
}