arXiv cs.CY· Leanne Tan, Keith Chua, Shaun Khoo, Onno P. Kampman, Lim Ruling Charmaine, Hazirah Hoosainsah, Kellie Yu Hui Sim, Kai He, Kenny Tsu Wei Choo, Roy Ka-Wei Lee·· 13 小时前AI 评分34
VEILED-SH：评估 AI 对间接自残与痛苦表达的响应基准
VEILED-SH: Vulnerability Expression through Indirect Language Evaluation Dataset (Self-Harm)
AI 导读
VEILED-SH 是一个评估 AI 对自杀、自残与痛苦间接表达单轮响应能力的基准及构建方法，包含 10,003 条提示词，覆盖咨询师聊天、论坛帖子、平台搜索与分诊场景，来源基于新加坡本地语料与公开数据集。
正文 · 原文
Abstract:AI systems may miss indirect, culturally situated distress or provide support mismatched to the application context. Existing benchmarks do not jointly address cultural variation, application context, and practitioner disagreement. We introduce VEILED-SH, a benchmark and construction methodology for evaluating single-turn responses to indirect expressions of suicide, self-harm, and distress. Its 10,003 prompts span counselor chat, forum posts, platform search, and triage, drawing on Singapore-grounded sources and public datasets. An expert-validated multi-label taxonomy distinguishes intent, communicative function, and response behavior. Six automated profiles, individually calibrated to three peer-support practitioners and three clinicians, retain separate reference policies for proxy-alignment scoring. Evaluating five systems on a coverage-enriched sample reveals variation in proxy alignment across profile-generated reference policies, even for the same system responses.
| Comments: | Content warning: this paper discusses suicide, self-harm, and distressing language |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY) |
| Cite as: | arXiv:2610.26190 [cs.CL] |
| (or arXiv:2610.26190v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.26190 arXiv-issued DOI via DataCite |
Submission history
From: Leanne Tan [view email]
[v1]
Mon, 19 Oct 2026 11:18:19 UTC (1,391 KB)
[v2]
Thu, 22 Oct 2026 04:39:04 UTC (1,391 KB)
来源：arXiv cs.CY · arxiv.org