RESCUE-BENCH: Towards Relation-Aware Multi-Party Emotional Support Conversation Systems
Summary
This paper introduces RESCUE-BENCH, a benchmark for evaluating large language models on relation-aware emotional support in multi-party conversations, revealing limitations in current models' ability to model interpersonal dynamics.
View Cached Full Text
Cached at: 09/10/26, 06:10 AM
Paper page - RESCUE-BENCH: Towards Relation-Aware Multi-Party Emotional Support Conversation Systems
Source: https://huggingface.co/papers/2609.09657
Abstract
This work introduces a benchmark for evaluating whether large language models can understand evolving interpersonal dynamics to provide effective emotional support in multi-party settings.
Existing emotional support conversation systems mainly focus on one-on-one seeker-supporter interactions and individual emotional states, leaving interpersonal relations in multi-party scenarios underexplored. In this work, we introducerelation-aware emotional support conversation, a new task that evaluates whether LLMs can capture and utilize the evolving dynamics of relationships to offer more effective emotional support. We constructRESCUE(Relation-aware Emotional Support ConversationUnderstanding and Evaluation Benchmark) from real couple and family interview conversations, containing 191 samples, 7,079 annotated turns, and 1,064.8 minutes of video. Based on rich annotations of socio-emotional and support-related dynamics,RESCUEdefines six tasks that evaluate two core capabilities required for relation-aware emotional support:Relational UnderstandingandRelation-Sensitive Support. Experiments with ten LLMs show that current models perform relatively well on tasks relying on local emotional or intervention cues, but struggle with relation-intensive tasks such asrelation pattern prediction,viewpoint prediction, andsupport strategy prediction. These findings reveal the limitations of current LLMs in modeling interpersonal relations and makingrelation-sensitive supportdecisions.
View arXiv pageView PDFGitHub1Add to collection
Get this paper in your agent:
hf papers read 2609\.09657
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2609.09657 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2609.09657 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2609.09657 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
ESCRAG-R1: Retrieval-Augmented Reinforcement Learning for Emotional Support Conversation
ESCRAG-R1 is a unified framework that integrates retrieval-based psychological guidance into Group Relative Policy Optimization to enhance emotional support conversations, mitigating superficial splicing and achieving natural integration of professional guidance and empathy.
RELATE: An Evaluation Framework for measuring Relational Orientation of Large Language Models
The paper introduces RELATE, a persona-conditioned evaluation framework measuring the relational orientation of large language models through inward-facing and outward-scaffolding language in multi-turn emotional support dialogues, finding that LLMs increasingly orient users toward themselves rather than real-world relationships over time.
DistressBench: Evaluating Multi-Dimensional Crisis Support in Large Language Models
DistressBench 提出一个由临床医生编写的 718 段对话基准，用于评估大模型在自杀与自残危机情境下的多维度支持能力，发现前沿模型平均只能满足 71.4% 的必要照护标准，且常在识别危机后未能引导用户寻求人工帮助。
τ-Emotion-Bench: Exposing the Interaction Fragility of LLM Agents through Real-World User Emotions
τ-Emotion-Bench introduces a ten-domain benchmark for evaluating tool-calling LLM agents under emotion-conditioned user interactions, alongside Tracer, a scalable agentic data-generation pipeline that improves Qwen3 models and generalizes to τ-Bench and τ-Trait, revealing that frontier agents remain unreliable when users behave emotionally.
SpeechEQ: Benchmarking Emotional Intelligence Quotient in Socially Aware Voice Conversational Models
SpeechEQ introduces a benchmark and dataset for evaluating emotional intelligence in speech-language models, covering 15 EQ subscales across 2,265 dialogues. Experiments reveal current models struggle with paralinguistic cues, exhibiting text-reliant shortcuts and other limitations.