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 roses were perfect, a vibrant crimson, arranged with effortless grace. They were delivered to her office. No note, just the flowers. The messenger just dropped them and vanished. Her carefully constructed facade of composure crumbled.

She felt a jolt of anger. Who would do this? Who would send such a beautiful, thoughtless gift? She felt the heat rising in her cheeks. It was as if she were a child, being deliberately ignored.

She slammed the flowers onto her desk, where they sat, a silent testament to something that felt akin to utter disdain. The rest of the day was shot. She avoided calls, and when someone approached, she made sharp, curt replies.

Emotion: insulted

Cluster: Anger / Hostility
PC1 (Valence): -0.98 Negative
PC2 (Disposition): 1.46

Role in Research

This story is one of 1,000 stories generated for the emotion insulted. 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 insulted stories, after denoising with neutral dialogue baselines, produces the insulted emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

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

Promoted:
l0.354
骂0.277
愤0.267
C0.237
🤬0.233
Suppressed:
own-0.325
давно-0.229
adventurous-0.223
soon-0.219
optimistic-0.217