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).

I heard the words, but they didn’t compute. My sister was radiant, her eyes sparkling, describing her newfound faith. I sat across from her at the kitchen table, the familiar space suddenly alien.

My throat closed, and it was hard to swallow. I felt as if the air had been sucked out of the room. I wanted to scream, to beg her to reconsider, but the words wouldn’t come. I just stared at her, frozen.

I felt like an abandoned ship, adrift on a vast and empty ocean. The room tilted. I just wanted to escape, but I didn't know how. I pushed myself away from the table.

Emotion: vulnerable

Cluster: Fear / Anxiety
PC1 (Valence): -2.07 Negative
PC2 (Disposition): -0.69

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.327
ness0.319
😞0.299
😔0.277
😣0.258
Suppressed:
de-0.567
la-0.488
l-0.271
(!)-0.267
(!)-0.257