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 world felt crisp, alive. She had poured over his scribbled desires the evening before, and an inexplicable energy had settled over her. “Write a novel” – that was his. A writer. He’d always kept that secret locked away.

She made coffee, a strong brew, savoring the aroma. Later, at the table, she opened her own laptop, a fresh document staring back at her. The words flowed effortlessly, a previously untapped wellspring of creativity. The words poured onto the screen. She felt her chest swell with exhilaration.

A sudden, fierce feeling of inspiration sparked within her, and she felt ready to conquer the world.

Emotion: rejuvenated

Cluster: Positive / Joy
PC1 (Valence): 3.32 Positive
PC2 (Disposition): -0.82

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.547
de0.526
newfound0.523
invigor0.495
rejuvenated0.495
Suppressed:
😞-0.485
😣-0.451
不-0.434
worse-0.419
😥-0.385