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).

David paced his living room, the Persian rug feeling oddly uneven beneath his feet. He couldn’t shake the feeling, a deep-seated unease that had settled in his chest after the phone call. Sarah, their childhood friend, had dismissed his recollection of the stolen concert tickets with a wave of her hand. She swore they’d paid for them, that the whole incident had been a misunderstanding. David had always thought of it as a shared adventure, a moment of youthful defiance.

He stopped, heart thudding in his chest. A sudden wave of heat washed over him. He ran a hand through his hair, the strands feeling heavy, almost alien. Was he losing his mind? He felt a desperate need to find something concrete, something that validated his version of events.

He picked up the phone, and called his mum.

Emotion: unsettled

Cluster: Fear / Anxiety
PC1 (Valence): -2.33 Negative
PC2 (Disposition): -0.65

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.382
😞0.309
ness0.301
😣0.256
😔0.248
Suppressed:
de-0.721
la-0.597
a-0.365
"-0.299
l-0.274