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).

Sarah practically skipped down the driveway, the keys to her new car jangling in her hand. The air vibrated with a palpable energy; the world seemed to hum along with her. She felt like she could leap over the fence and dance with the squirrels. A shadow fell across her path.

It was Mr. Henderson, his face creased with a troubled frown. โ€œSarah, dear,โ€ he said, pointing towards the sprawling maple tree that straddled their properties. "I was wondering if we could talk about having it removed."

Her smile widened, her teeth gleaming. She took a deep breath, the crisp air filling her lungs and expanding her chest. Her thoughts, normally a jumble, were now clear as a bell. "Of course, Mr. Henderson! What did you have in mind?"

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
๐Ÿ˜ž-0.470
๐Ÿ˜ฃ-0.337
L-0.334