# Expert-Level Crisis Detection in Mental Health Conversations

arXiv: https://arxiv.org/abs/2606.10380 · HTML: https://mlsanity.com/paper/2606.10380/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 32 references that are in the map.
- First posted on arXiv 9 Jun 2026.

## Builds on (most cited first, 13 of 13)

- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [Cactus: Towards Psychological Counseling Conversations using Cognitive Behavioral Theory](https://mlsanity.com/paper/2407.03103/) (2024)
- [HealMe: Harnessing Cognitive Reframing in Large Language Models for Psychotherapy](https://mlsanity.com/paper/2403.05574/) (2024)
- [CBT-LLM: A Chinese Large Language Model for Cognitive Behavioral Therapy-based Mental Health Question Answering](https://mlsanity.com/paper/2403.16008/) (2024)
- [MentalChat16K: A Benchmark Dataset for Conversational Mental Health Assistance](https://mlsanity.com/paper/2503.13509/) (2025)
- [Can Large Language Models Identify Implicit Suicidal Ideation? An Empirical Evaluation](https://mlsanity.com/paper/2502.17899/) (2025)
- [Rethinking the Alignment of Psychotherapy Dialogue Generation with Motivational Interviewing Strategies](https://mlsanity.com/paper/2408.06527/) (2024)
- [Trustworthy AI Psychotherapy: Multi-Agent LLM Workflow for Counseling and Explainable Mental Disorder Diagnosis](https://mlsanity.com/paper/2508.11398/) (2025)
- [CRADLE Bench: A Clinician-Annotated Benchmark for Multi-Faceted Mental Health Crisis and Safety Risk Detection](https://mlsanity.com/paper/2510.23845/) (2025)
- [Just a Scratch: Enhancing LLM Capabilities for Self-harm Detection through Intent Differentiation and Emoji Interpretation](https://mlsanity.com/paper/2506.05073/) (2025)
- [LOST: A Mental Health Dataset of Low Self-esteem in Reddit Posts](https://mlsanity.com/paper/2306.05596/) (2023)
- [A Layered Multi-Expert Framework for Long-Context Mental Health Assessments](https://mlsanity.com/paper/2501.13951/) (2025)
- [Roleplaying with Structure: Synthetic Therapist-Client Conversation Generation from Questionnaires](https://mlsanity.com/paper/2510.25384/) (2025)

## Abstract (arXiv)

Real-world crisis intervention is inherently conversational, yet existing research largely focuses on static texts. When applied to multi-turn dialogues, current models exhibit significant performance degradation, struggling to track risk signals that emerge as context evolves. To address this gap, we introduce CRADLE-Dialogue, a clinician-annotated benchmark for turn-level crisis detection in conversational settings. The dataset features 600 dialogues with multi-label annotations across clinically grounded risks, including suicide ideation, self-harm, and child abuse, distinguishing past from ongoing risk. We further propose an Alert-Confirm evaluation protocol that distinguishes early warning signals (Alert) from turns where a specific crisis becomes explicitly identifiable (Confirm), reflecting the clinical need to intervene before risk becomes explicit. Experiments show that identifying when risk emerges is much harder than recognizing that it exists: models achieve only mid-40% to high-60% Micro F1. Additionally, we release a synthetic training corpus and a 32B-parameter model that substantially outperforms existing open-source models and achieves competitive or superior results against proprietary models across turn-level, dialogue-level, and confirm-only evaluation settings.