A linear emotional direction drives VLM emotion expression

Zhang, Sun, Xie & Tian extract a layer-specific linear emotional direction in Qwen3-VL-4B-Instruct, Qwen3-VL-8B-Instruct, and LLaVA-OneVision-1.5-4B-Instruct by averaging contrastive emotional-vs-neutral hidden-state differences [zhang-etal-2026-interpreting-and-enhancing-emotional-circuits-in-large-vision-language-models] Activation steering with this direction causally raises the model's emotion-expression hit rate [zhang-etal-2026-interpreting-and-enhancing-emotional-circuits-in-large-vision-language-models] Backward activation patching from the direction localizes the upstream attention heads and MLP neurons that construct it [zhang-etal-2026-interpreting-and-enhancing-emotional-circuits-in-large-vision-language-models] The training-free VEENA framework built on the direction enhances or suppresses emotional expression at inference without retraining [zhang-etal-2026-interpreting-and-enhancing-emotional-circuits-in-large-vision-language-models]