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 first thing David did when he woke up was open his curtains, greeting the morning sun. The sight of the half-renovated house next door, the scaffolding casting long shadows on the lawn, was a welcome sight. He felt a thrill of excitement, an energy that propelled him out of bed.

He began the day with a jog, pushing himself harder than usual. The pounding of his feet on the pavement seemed synchronized with the hammering echoing from next door. He had his own dream project: transforming his cramped attic into a home office. He could finally start writing the novel.

He returned home, a spring in his step, and immediately sat down at his computer.

Emotion: hopeful

Cluster: Positive / Joy
PC1 (Valence): 3.68 Positive
PC2 (Disposition): -0.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.715
la0.549
B0.468
🀩0.452
enthusi0.433
Suppressed:
S-0.603
不-0.401
😞-0.399
worse-0.394
C-0.386