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).

Maya swung her legs, kicking at the worn armchair in her grandmother's attic. Dust motes swirled in the slant of the afternoon sun, illuminating the forgotten treasures. She felt a lightness, a buoyant feeling that threatened to lift her right off the floor. The air itself seemed charged with a happy energy. She'd just been accepted to her dream school. Then, another email.

Her hand shot out, snatching her phone from the rickety side table. A notification from the university. Her thumb stammered as she clicked the link. The university's website. Her stomach did a little flip, a playful little somersault. There it was, bold as brass: "Essay Examples: Overcoming Adversity." Her essay was right there on the screen. The words she’d written, the story she’d shared, were now public. A giggle escaped her lips.

Emotion: vibrant

Cluster: Positive / Joy
PC1 (Valence): 2.83 Positive
PC2 (Disposition): -1.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.518
!0.459
!".0.436
de0.434
!).0.409
Suppressed:
S-0.890
L-0.478
P-0.407
H-0.365
😞-0.306