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
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add Causal-ERC (fusion plus causal prompting) over the modified Emotion-LLaMA baseline | Llama-3.1-8B-Instruct LoRA · IEMOCAP | W-F1 (%): 69.06 -> 71.89 | Improved |
| add Causal-ERC over the baseline | Llama-3.1-8B-Instruct LoRA · MELD | W-F1 (%): 67.84 -> 70.25 | Improved |
| add causal prompting (ablation) | Causal-ERC · IEMOCAP | W-F1 (%): 70.68 -> 71.89 | Improved |
| add causal prompting (ablation) | Causal-ERC · MELD | W-F1 (%): 68.82 -> 70.25 | Improved |
| add multimodal fusion module (ablation) | Causal-ERC · IEMOCAP | W-F1 (%): 70.14 -> 71.89 | Improved |
| add multimodal fusion module (ablation) | Causal-ERC · MELD | W-F1 (%): 68.38 -> 70.25 | Improved |
| add Soft-HGR consistency loss (ablation) | Causal-ERC · IEMOCAP | W-F1 (%): 71.01 -> 71.89 | Improved |
| add consistency loss (ablation) | Causal-ERC · MELD | W-F1 (%): 69.37 -> 70.25 | Improved |
| add audio and visual modalities to text (ablation) | Causal-ERC · IEMOCAP | W-F1 (%): 69.13 -> 71.89 | Improved |
| add audio and visual modalities to text (ablation) | Causal-ERC · MELD | W-F1 (%): 68.10 -> 70.25 | Improved |
| increase history window size (ablation) | Causal-ERC · IEMOCAP validation | W-F1 improvement over baseline | Mixed |
| add causal prompting (ablation) | Causal-ERC · MELD per-class | Disgust F1 (%): Surprise 64.93 -> 62.96, Disgust 51.09 -> 45.53 | Mixed |
Closest papers by meaning
- Adaptive Graph Learning for Multimodal Conversational Emotion Detection (AAAI 2024 · similarity 0.88)
- Emotion-LLaMA: Multimodal Emotion Recognition and Reasoning with Instruction Tuning (NeurIPS 2024 · similarity 0.88)
- AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models (ICML 2025 · similarity 0.87)
- Utterance-level Emotion Recognition in Conversation with Conversation-level Supervision (AAAI 2025 · similarity 0.87)
- Debiased Multimodal Understanding for Human Language Sequences (AAAI 2025 · similarity 0.87)
- Catch Your Emotion: Sharpening Emotion Perception in Multimodal Large Language Models (ICML 2025 · similarity 0.87)
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.
