AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Causal-ERC: A Multimodal Framework with Causal Prompting for Emotion Recognition in Conversations with Large Language Models

Ran Jing, Geng Tu, Yice Zhang, Ruifeng Xu

AAAI 2026 · Large language models › Language, safety and interpretability · approach: Faithfulness of LLM explanations · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Causal-ERC (fusion plus causal prompting) over the modified Emotion-LLaMA baselineLlama-3.1-8B-Instruct LoRA · IEMOCAPW-F1 (%): 69.06 -> 71.89Improved
add Causal-ERC over the baselineLlama-3.1-8B-Instruct LoRA · MELDW-F1 (%): 67.84 -> 70.25Improved
add causal prompting (ablation)Causal-ERC · IEMOCAPW-F1 (%): 70.68 -> 71.89Improved
add causal prompting (ablation)Causal-ERC · MELDW-F1 (%): 68.82 -> 70.25Improved
add multimodal fusion module (ablation)Causal-ERC · IEMOCAPW-F1 (%): 70.14 -> 71.89Improved
add multimodal fusion module (ablation)Causal-ERC · MELDW-F1 (%): 68.38 -> 70.25Improved
add Soft-HGR consistency loss (ablation)Causal-ERC · IEMOCAPW-F1 (%): 71.01 -> 71.89Improved
add consistency loss (ablation)Causal-ERC · MELDW-F1 (%): 69.37 -> 70.25Improved
add audio and visual modalities to text (ablation)Causal-ERC · IEMOCAPW-F1 (%): 69.13 -> 71.89Improved
add audio and visual modalities to text (ablation)Causal-ERC · MELDW-F1 (%): 68.10 -> 70.25Improved
increase history window size (ablation)Causal-ERC · IEMOCAP validationW-F1 improvement over baselineMixed
add causal prompting (ablation)Causal-ERC · MELD per-classDisgust F1 (%): Surprise 64.93 -> 62.96, Disgust 51.09 -> 45.53Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy