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).

“Are you even ready yet?” Daniel’s sister called from the hallway. He stared blankly at his closet, its contents a vibrant tapestry of clothes that suddenly felt meaningless. Today was Maria’s birthday, a rooftop bar with a DJ. He needed to be dressed. He needed to be happy.

The silence in the room stretched and became heavy. He remembered the text message he'd received from Michael, a quiet dinner at his house. The thought of a cozy evening sounded appealing, but he wasn’t up to making the choice.

He ran a hand through his hair, feeling the strands catch on his fingers. He hadn't shaved in days. He felt a deep sense of apathy, a profound lack of interest in everything. The world seemed to have lost its color.

Emotion: melancholy

Cluster: Sadness / Despair
PC1 (Valence): -1.78 Negative
PC2 (Disposition): -1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.445
😞0.376
😔0.337
いつも0.327
sadness0.312
Suppressed:
de-0.739
la-0.548
l-0.497
/-0.314
!-0.290