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 attic air hung thick and dusty, catching in Agnes’ throat. She coughed, shoving aside a moth-eaten tapestry to reach the trunk. Her grandmother's belongings, a repository of carefully curated history, a history she was beginning to doubt. Why hadn't she just let it go?

Her fingers fumbled with the rusted lock. Inside, a jumble of faded photographs. She snatched one up. A laughing woman, her grandmother, arm-in-arm with a man who was most certainly not her grandfather. A man with a devilish grin and eyes that danced with mirth.

A vein throbbed in Agnes' temple. The family tale: a whirlwind romance, a love that endured a lifetime. Lies. All lies. She crumpled the photograph, then, immediately, smoothed it out again. She couldn't believe it.

Emotion: irate

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.45

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.430
🤬0.350
aggravated0.343
骂0.340
恨0.331
Suppressed:
H-0.292
soon-0.289
☺️-0.289
optimistic-0.258
latter-0.253