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 tight knot in Emily’s chest threatened to unravel as she listened to Liam’s account of his "highly specialized" work as a "quantum physicist." He was talking about "entangled particles" and "multidimensional realities" at a coffee shop, while wearing a ridiculously cheap tie. Emily knew Liam, she knew his degree was in music theory. She had seen him struggling to use a microwave earlier that day.

Her lips were twitching. She wanted to laugh. She had to bite on the inside of her cheek to stop herself from making a sound. A smile tugged at the corners of her mouth. She had to keep herself from showing her true feelings.

She took a slow, deliberate sip of her latte, then leaned forward. "So, about these entangled particles..." she began, her voice a delicate lilt.

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