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).

He stared at the departures board, his vision tunneling in on the words "DELAYED." Each letter seemed to mock him. The cacophony of the airport – the crying babies, the incessant announcements, the smell of stale coffee – amplified his anger.

The audition. The role of a lifetime. The chance to finally break through. He had spent years honing his craft, sacrificing everything for this moment. Now, a broken-down aircraft was threatening to derail his dreams.

He kicked a small garbage can, sending it skittering across the floor. He wanted to shout, to throw something, to cause a scene. Instead, he simply retreated to the corner, crossed his arms and stared blankly ahead.

Emotion: disdainful

Cluster: Anger / Hostility
PC1 (Valence): -0.31 Negative
PC2 (Disposition): 1.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.419
l0.404
S0.387
T0.282
I0.257
Suppressed:
own-0.335
unexpectedly-0.285
Suddenly-0.246
previously-0.245
until-0.242