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 security cameras, the gleaming chrome reflecting the midday sun. His neighbor, Frank, the guy who always mowed his own lawn and wore old t-shirts, had a state-of-the-art security system. A *very* state-of-the-art security system.

Mark had spent years thinking of Frank as a bit of a slob, making snide remarks about his rusty car and his habit of leaving the trash cans out. Now, looking at those cameras, he felt a sudden and overwhelming sense of self-consciousness.

He thought of all the times he'd complained about Frank's "loud" lawnmower. He thought about the time he'd let his dog, off leash, on Frank's lawn.

Emotion: mortified

Cluster: Embarrassment
PC1 (Valence): -1.36 Negative
PC2 (Disposition): 0.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.406
尴尬0.257
C0.252
worse0.245
逃0.242
Suppressed:
own-0.340
了他-0.218
他对-0.203
ል-0.194
楽し-0.193