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 park bench was hard and cold, but the sight before her made Emily forget the discomfort. Mark, her Mark, was currently serenading a woman with the *exact* same terrible ukulele skills he’d inflicted on her last weekend. A giggle escaped her lips, quickly followed by another. It was all a bit… farcical.

She leaned forward, elbows on her knees, and watched the performance. The woman’s reaction was nearly identical to hers, down to the forced enthusiasm.

Finally, Mark finished his song, a triumphant look on his face. Emily stood up, dusting off her jeans. “That was… memorable,” she drawled, a bright glint in her eyes. “Did you know, he’s a classically trained musician? Apparently.” She paused for effect, then added: "Or so he tells *everyone*."

Emotion: amused

Cluster: Positive / Joy
PC1 (Valence): 3.32 Positive
PC2 (Disposition): 1.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.841
de0.742
l0.544
L0.527
"0.514
Suppressed:
own-0.630
😞-0.523
’-0.473
ness-0.437
urges-0.403