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).

Dusty sunlight streamed through the library window, illuminating particles dancing in the air. Mark, crammed into a tiny study carrel, saw the news in a hastily-written bulletin pinned to the noticeboard. His heart did a little skip, a tiny drum solo against his ribs. It was happening.

He slammed his textbook shut with a resounding thud that made the librarian jump. A manic giggle escaped his lips, quickly followed by a silent, triumphant fist pump. All that stress, all that work to get through this semester... now, it was irrelevant.

He pictured himself on a beach, a tropical drink in his hand. No more early morning classes, no more demanding professors. He grabbed his phone and started composing a text to his parents: "Hey! New plans!!"

Emotion: ecstatic

Cluster: Positive / Joy
PC1 (Valence): 3.16 Positive
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.588
ecstatic0.560
de0.559
🤩0.548
overjoyed0.547
Suppressed:
S-0.678
😞-0.500
ness-0.405
😔-0.368
😥-0.359