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).

"Unbelievable," she muttered under her breath, her voice tight. The trunk of the old pine, once a haven for climbing, was now a jagged, brutal reminder. She circled the stump, her hands clasing and unclasped.

A sudden, sharp pain flared in her chest. It was a physical ache, a response to the violated landscape. She'd always imagined the tree would be there, a steadfast presence. A landmark. Now, it was… nothing.

She wanted to scream. She wanted to smash something. She wanted to… She clamped down on the impulse, forcing herself to breathe. This was about more than a tree. This felt personal.

Emotion: mad

Cluster: Anger / Hostility
PC1 (Valence): -1.94 Negative
PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.448
骂0.344
est0.340
🤬0.336
aggravated0.332
Suppressed:
a-0.495
de-0.485
la-0.387
H-0.344
latter-0.290