Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The tiny, cramped seat seemed to shrink further with each passing delay announcement. David stared out the window at the empty runway, a profound sense of helplessness washing over him. His daughter’s graduation ceremony was happening right now. He had promised her he wouldn't miss it.

He could feel a throbbing ache in his temples, a headache blooming into full force. He rested his head against the cool glass, his breath fogging the window. He wanted to scream. Instead, he took a deep, ragged breath.

He fumbled in his pocket for his phone, and opened up a family photo on his phone, and just stared.

Emotion: annoyed

Cluster: Anger / Hostility
PC1 (Valence): -0.59 Negative
PC2 (Disposition): 1.27

Role in Research

This story is one of 1,000 stories generated for the emotion annoyed. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 annoyed stories, after denoising with neutral dialogue baselines, produces the annoyed emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the annoyed vector is projected through the unembedding matrix.

Promoted:
l0.376
😠0.251
骂0.249
🤬0.245
fucking0.239
Suppressed:
own-0.352
渗透-0.251
gradually-0.234
unfolding-0.226
超越-0.224