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 crayon slipped from Maya’s numb fingers. She stared, mouth agape, at the picture her niece, Lily, had just presented. It depicted a lanky figure with mismatched socks and a mischievous grin, exactly as Maya had remembered “Zip” from her own childhood. He was perched on a swing set, just like Maya used to envision.

Her heart hammered against her ribs. "Lily," she stammered, voice tight, "who is this?"

Lily, oblivious, pointed a grubby finger. "It's Ziggy! He comes to play with me in the garden. He's very silly." Maya felt a cold wave wash over her, a prickling sensation crawling up her spine. This couldn't be happening.

Emotion: astonished

Cluster: Surprise / Confusion
PC1 (Valence): -0.84 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.448
de0.350
la0.323
l0.307
incred0.266
Suppressed:
own-0.289
cheerfully-0.231
enjoyed-0.224
开心-0.221
ี่ยว-0.219