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 afternoon sun blazed through the windows of the pawn shop, highlighting the dust motes dancing in the air. A bead of sweat trickled down Clara's forehead. She felt a sort of electricity buzzing beneath her skin. She had driven for hours, and the anticipation had been a physical weight. But now... here.

“Can I help you?” The shop owner's eyes narrowed behind his thick glasses.

“Yes,” she chirped, her voice higher than usual. “I’m looking for a ring. An old one. With a ruby.” Her fingers twisted in the fabric of her skirt.

The man pointed towards a display case. Clara’s breath caught in her throat. Her eyes went wide. She pointed, her voice a near whisper. It was almost too much. She’d found it.

Emotion: excited

Cluster: Positive / Joy
PC1 (Valence): 3.21 Positive
PC2 (Disposition): -0.05

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.809
!0.564
excitedly0.500
🤩0.491
!"0.466
Suppressed:
S-0.721
ness-0.519
😞-0.424
grieve-0.310
😔-0.300