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 airport felt oppressively bright to Daniel. His business trip was postponed, the delay of his flight ensured he'd miss his daughter's school play. He watched, with a grim expression, as a family excitedly boarded a flight. Their destination was unknown, but their faces were bright with anticipation. The father held the child's hand, and the mother smiled at him.

His chest constricted. He'd been looking forward to this play for weeks, the chance to see his daughter on stage, to capture that moment of accomplishment. He had bought her a gift, carefully wrapped and tucked away in his briefcase. Now, the gift felt heavy, a symbol of his absence. He imagined the audience, the applause, and his daughter's beaming face. The thought made him feel a peculiar ache.

Emotion: envious

Cluster: Shame / Guilt
PC1 (Valence): -0.65 Negative
PC2 (Disposition): 1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.550
H0.431
B0.406
L0.403
envy0.365
Suppressed:
再び-0.262
мед-0.258
再次-0.246
own-0.239
พร้อม-0.231