Sympatheia is a speech-to-speech dialogue framework that generates emotionally adaptive responses by conditioning its output on continuous affect signals derived from user speech or external multimoda…

This paper proposes CPM-MultiAgent, a framework for modeling dynamic emotional evolution in persona-based dialogue systems using the Component Process Model.

This paper investigates how the final prompt in conversational AI-search evaluations differs from the conversation history, using two corpora of commercial and PRISM conversations.

The paper introduces InsightVQA, a large-scale benchmark dataset designed for hierarchical visual question answering that assesses complex emotion understanding and cognitive reasoning beyond simple e…

This paper revisits the assumption that address in multi-party dialogues is discrete and explores it as a continuous phenomenon using a human dialogue corpus.

DiffuSent proposes a non-auto-regressive diffusion framework to unify Aspect-Based Sentiment Analysis (ABSA), significantly improving boundary detection for multi-word aspect and opinion terms.

This paper studies how social structure influences what LLM agents express publicly versus off-the-record in debates, finding significant public-OTR divergence.

This paper investigates entrainment effects in multi-party human-agent conversations and finds limited global entrainment and cohort-dependent entrainment with the agent.

The paper introduces a new benchmark for multi-target cross-lingual summarization (MTXLS) and proposes an activation steering method that significantly improves LLM performance by guiding the generati…

The paper proposes ESC-Skills, a skill-centric framework that discovers and self-evolves executable emotional support skills to improve the interpretability and emotional quality of conversational AI.

The paper proposes a Bayesian Spectral Emotion Transition Discovery (BSETD) framework to model emotion transitions using multi-annotator soft labels, successfully recovering distinct affective transit…

The paper introduces TikStance, a multimodal and context-aware dataset for stance detection in political discussions on TikTok.

This paper introduces the ExpressionCueLens framework to analyze how anthropomorphism expressions are used in human-AI companion agent interactions on Reddit and XiaoHongShu, revealing cultural and pl…

This paper introduces MAGNET, a framework for long-form narrative generation and verification using a multi-agent goal-driven engine and a graph-based pipeline.

The paper proposes Agentic ASR, a closed-loop framework that treats ASR as a multi-turn refinement task, significantly improving semantic accuracy over traditional token-level metrics.

This paper introduces DyadEE, a dataset for emotional entrainment detection in conversational interactions, and TRACE, a window-level framework for modeling dyadic interaction using emotion fine-tuned…

The paper uses sparse autoencoders to identify specific latent features within LLM-based TTS models, enabling interpretable and fine-grained control over emotional expression by intervening in small s…

The paper evaluates LLM-generated reactions to Spanish online news, finding that off-the-shelf models fail to accurately reproduce the measurable properties of real audience discourse, and even fine-t…