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).

From the moment he woke, Mateo felt like he could run a marathon. He devoured his breakfast, leaving the plate completely clean. The day pulsed with possibilities. He had dreamt of this scholarship. Now, it was finally happening! The financial aid office was only a short walk away.

The atmosphere in the financial aid office seemed to thicken the moment he stepped inside. The air felt heavy, like a humid summer day. Ms. Chen, the director, beckoned him into her office with a serious expression. Mateo’s smile began to wane. He felt his stomach do a flip, a sensation like freefalling in an elevator.

The mistake was revealed. A clerical error. His last name was very similar. The money was for someone else. Mateo felt a strange sort of heat surge through him. The good news that he had known for a week was a lie. He tried to speak, but the words wouldn't come. "So, I just go back home and find a job?" he asked in a monotone.

Emotion: vibrant

Cluster: Positive / Joy
PC1 (Valence): 2.83 Positive
PC2 (Disposition): -1.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.518
!0.459
!".0.436
de0.434
!).0.409
Suppressed:
S-0.890
L-0.478
P-0.407
H-0.365
😞-0.306