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).

As the door swung open, the scent of fresh paint and the buzz of construction filled her senses. Emily’s transfer to the Project Management team was a shock, and the current office renovations added to her discomfort. Her fingers drummed against her leg. She had to learn everything all over again.

She took a shaky step inside. The new office was still a work in progress, with exposed wires and scattered tools. The noise was constant, a combination of hammering, sawing, and the chatter of construction workers. It made her skin prickle.

Her new supervisor, a woman in a hard hat, gave her a curt nod. "Welcome. We're behind schedule, so let's get you up to speed."

Emotion: nervous

Cluster: Fear / Anxiety
PC1 (Valence): -1.08 Negative
PC2 (Disposition): -0.47

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.229
trem0.203
😰0.203
😨0.201
緊張0.200
Suppressed:
la-0.397
que-0.260
a-0.242
B-0.217
H-0.210