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 restaurant was too loud. The music was too loud. The waiter was too cheerful. Sarah pushed the untouched plate of food away, the texture of the sauce triggering a sudden, unwelcome wave of nausea. She wanted to scream.

The condition had taken away so much: her energy, her appetite, her freedom. The specialist’s recommendations, the endless tests, the medication side effects…it was a prison. Finding someone else with the same condition, a person to share her burden? The idea was repulsive.

Yet, a message had come through, a number she couldn't bring herself to delete. She almost deleted it, then deleted the draft message. She turned off her phone.

Emotion: angry

Cluster: Anger / Hostility
PC1 (Valence): -2.05 Negative
PC2 (Disposition): 1.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.354
est0.354
aggravated0.342
🤬0.337
恨0.323
Suppressed:
de-0.424
a-0.411
la-0.341
H-0.320
L-0.296