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 champagne flute felt weightless in Amelia’s hand as she twirled on the dance floor, her laughter echoing above the band. The sequined dress swirled around her, a kaleidoscope of color in the warm lighting of the ballroom. She felt like she could fly.

Then, she saw him. Liam. Standing near the bar, looking as handsome and a little awkward as she remembered. Her smile faltered for a fraction of a second, but then, she lifted her chin and winked at the bridesmaid she was dancing with.

“He looks like he’s seen a ghost,” Amelia whispered, her voice bubbling with amusement. She took another sip of champagne and, without missing a beat, started singing along to the cheesy pop song, her voice carrying across the room, full and bright.

Emotion: joyful

Cluster: Positive / Joy
PC1 (Valence): 3.96 Positive
PC2 (Disposition): -0.67

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.615
de0.600
!0.572
excitedly0.552
enthusi0.533
Suppressed:
S-0.893
ness-0.513
😞-0.470
😣-0.337
L-0.334