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 poplar, which had stood so tall and proud, was now resting on Mr. Henderson's shed. My eyes widened with a slight feeling of… delight. I covered my mouth with a hand, suppressing a giggle that threatened to burst forth. The contrast between the tree's former majesty and its current state was just too much.

I picked up my phone, quickly making a call. “You’ll never guess what happened,” I began, my voice already laced with mirth. The thought of sharing the news and hearing the reactions of my friends and family was so much fun. I stretched, a delightful energy coursing through me.

This was going to be the talk of the neighborhood.

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