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).

Sunlight streamed through the window, warming Amelia's face. She stretched, a long, luxurious yawn escaping her lips. The air smelled of freshly brewed coffee, a comforting aroma that always made her feel… buoyant. Today was the day. The interview for the marketing assistant role at "Bloom & Thrive" was finally happening. She glanced at the neatly stacked resume and portfolio. A small smile played on her lips. She felt absolutely ready.

The drive to the office was a breeze. She put on her favorite upbeat playlist and tapped her fingers on the steering wheel, completely uninhibited. She found herself humming along to the music, a lightness in her chest.

During the interview, Amelia found herself speaking with an unusual confidence. She answered each question with clarity and a genuine enthusiasm that surprised even herself. As she left the building, she practically bounced on her way back to the car.

Emotion: refreshed

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.998
de0.703
!).0.384
(!)0.381
!".0.379
Suppressed:
S-0.782
C-0.478
L-0.414
😣-0.388
😞-0.376