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).

She slammed the biography onto the table in front of the lawyer. The man, a well-meaning but often oblivious type, blinked, then adjusted his glasses. "Well, this is certainly… a surprise," he stammered. Her lips curled slightly as she watched him. The book felt like a violation.

The lawyer flipped through the glossy pages, his eyebrows rising at certain passages. She could practically feel the writer’s false understanding of her, the way he’d twisted her life into a neat, easily digestible narrative. She tapped a finger impatiently on the table.

"I need this stopped," she said, her voice low and dangerous. "Every single copy. I want a retraction. I want legal action. And I want it all done yesterday."

Emotion: scornful

Cluster: Anger / Hostility
PC1 (Valence): -0.76 Negative
PC2 (Disposition): 2.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.406
C0.391
S0.276
I0.268
愤0.243
Suppressed:
own-0.396
unexpectedly-0.288
Suddenly-0.260
soon-0.256
serendip-0.256