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 screen of her laptop reflected Sarah's strained expression back at her. She stared at the spreadsheet, her work, her life, suddenly seeming meaningless. The discovery that Michael, her "CEO" boyfriend, was actually a telemarketer, had left her feeling…numb.

She had built her whole life around his supposed success, her ambitions, her social circle. She'd invested her emotions, her dreams. Now, everything felt like a lie. She wanted to scream, to shatter the illusion that surrounded her.

She closed her laptop with a snap, the sound echoing in the silent room. She needed to be far away, somewhere new. The walls of her apartment felt like they were closing in. Her head began to ache. She felt suffocated by the deception.

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