Affective Flow Language Model for Emotional Support Conversation

TL;DR

Affective Flow Language Model (AFlow) enhances emotional support conversations by modeling continuous affective trajectories, improving strategy coherence and response quality over existing methods.

Contribution

Introduces AFlow, a framework with fine-grained supervision on dialogue prefixes that models affective flow for better multi-turn emotional support conversations.

Findings

AFlow outperforms competitive baselines in diverse emotional contexts.

AFlow with a compact backbone surpasses proprietary LMMs like GPT-4o and Claude-3.5.

The code is publicly available at https://github.com/chz2025/AffectiveFlow.

Abstract

Large language models (LLMs) have been widely applied to emotional support conversation (ESC). However, complex multi-turn support remains challenging.This is because existing alignment schemes rely on sparse outcome-level signals, thus offering limited supervision for intermediate strategy decisions. To fill this gap, this paper proposes affective flow language model for emotional support conversation (AFlow), a framework that introduces fine-grained supervision on dialogue prefixes by modeling a continuous affective flow along multi-turn trajectories. AFlow can estimate intermediate utility over searched trajectories and learn preference-consistent strategy transitions. To improve strategy coherence and empathetic response quality, a subpath-level flow-balance objective is presented to propagate preference signals to intermediate states. Experiment results show consistent and significant improvements over competitive baselines in diverse emotional contexts. Remarkably, AFlow with a compact open-source backbone outperforms proprietary LMMs such as GPT-4o and Claude-3.5 on major ESC metrics. Our code is available at https://github.com/chz2025/AffectiveFlow.