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 never consented to this!" Amelia hissed, her voice barely a whisper, as she confronted the publisher. The polished glass of the building reflected her distorted image. Her hands trembled as she clutched the book, its cover featuring a picture of her beaming. Inside, it depicted a life of constant triumph and effortless grace. In reality, her days felt like an uphill climb, each step a struggle. Now, everyone would read about a person who didn't exist. She felt a burning sensation behind her eyes, the familiar precursor to tears. She didn't want to break down in front of them, but this was a betrayal she hadn't anticipated.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236