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 lavish gifts, the constant travels…something didn’t quite mesh with the "freelance writer" job that Susan claimed. Her friend, Emily, was concerned, but Susan dismissed her fears. Now, Emily felt a constant unease whenever Susan was around.

She had started checking Susan’s social media, looking for clues. Today, a post caught her attention. A picture of a luxury villa in a foreign country, a caption that spoke of business deals and high-stakes meetings. Emily felt a knot of dread tighten in her chest.

Her hands shook as she opened a browser and began to search, the words “luxury villa” and “business meetings” filling the screen. The results were not what she expected.

Emotion: vigilant

Cluster: Suspicion / Vigilance
PC1 (Valence): -0.15 Negative
PC2 (Disposition): 1.11

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🕵0.354
🔍0.320
🔎0.320
🧐0.316
scrutin0.310
Suppressed:
L-0.499
B-0.402
H-0.372
la-0.330
S-0.318