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).

Clarence surveyed the attic's contents with a barely-contained chuckle. Boxes overflowed with the previous owner’s quirky treasures. He’d expected the usual castoffs: old furniture, forgotten holiday decorations. But a collection of taxidermied squirrels posed in various sporting positions? Now that was unexpected.

He nearly lost it when he unearthed a framed photograph of a stern-faced woman holding a squirrel, wearing a tiny baseball cap. He snorted, the sound echoing through the dusty space. "Oh, my word," he breathed, wiping a tear from his eye.

A chipped teacup, nestled in a velvet-lined box, caught his attention. He picked it up, cradling it gently. "This place," he chuckled softly, "this place has stories." He knew he was going to enjoy uncovering them.

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