TIER: Threat Implicitness Benchmark for Evaluating LLM Safety Behaviors

Abstract

Current LLM safety benchmarks largely rely on binary metrics, overlooking how models respond to harmful prompts with varying threat implicitness. We introduce TIER, a Threat Implicitness Benchmark for behavioral safety evaluation of LLMs. TIER covers four risk domains and four threat levels, from explicit harmful requests to sophisticated jailbreaks. Responses are assessed using a six-label behavior scale and two independent LLM judges. Experiments on six open-weight LLMs show that safety behavi

Categories

Framework mappings

Suggested from the entry's categories.

@misc{trinhthi2026tier,
  title = {{TIER: Threat Implicitness Benchmark for Evaluating LLM Safety Behaviors}},
  author = {Thu-Hien Trinh-Thi and Hai-Yen Vong and Thanh-Ha Ung-Dung and Tram Ho},
  year = {2026},
  month = sep,
  eprint = {2609.05117},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.05117}
}