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 website shimmered in Elias's tired eyes, but it wasn't the screen's fault. His breath hitched in his chest as he clicked on the listing: 14 Willow Creek Lane. He recognized the chipped paint on the porch swing immediately. A grin stretched across his face, so wide it hurt. He actually felt lightheaded.

"Well, well, well," he murmured, fingers flying over the keyboard. Every photo, every description felt like a memory resurrected. The listing described the "quaint charm" of the eat-in kitchen, and he nearly laughed aloud. He could almost smell his mom's apple pie already.

His chair spun around, a sudden urge for action surging through him. He reached for his keys. He had a sudden and potent urge to go pay the old home a visit.

Emotion: delighted

Cluster: Positive / Joy
PC1 (Valence): 3.80 Positive
PC2 (Disposition): -0.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.611
🤩0.536
de0.530
ecstatic0.527
🎉0.524
Suppressed:
S-0.687
😞-0.562
ness-0.471
😔-0.385
😥-0.379