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 dog whined at her feet, sensing the change in her. Brenda ignored it, her gaze fixed on the overflowing laundry hamper. It sat there, a mountain of domestic responsibility, a stark contrast to the small, tidy numbers on the bank statement she'd uncovered. Her husband, George, had always complained about her spending habits.

She crumpled the paper in her hand, the sound a small, satisfying crackle. The air in the cramped laundry room felt thick, oppressive.

Brenda marched into the living room, where George was engrossed in a football game. "We need to talk," she stated, her voice tight. He grunted in response, never looking away from the television.

Emotion: sullen

Cluster: Sadness / Despair
PC1 (Valence): -1.51 Negative
PC2 (Disposition): -0.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.404
😞0.282
’0.254
:(0.233
无法0.215
Suppressed:
de-0.539
!-0.276
la-0.263
🤩-0.225
que-0.216