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 afternoon sun streamed through the window, illuminating the dust motes dancing in the air. Ms. Jones, after a long and grueling day, was in a particularly grumpy mood. She slumped into her chair, ready to just head home. Then, the office clerk popped her head in, holding a small bouquet of wildflowers.

They were delivered by a former student, a shy girl named Lily. A card simply said "Thank you, Ms. Jones." She brought the flowers to her nose and inhaled deeply. The fresh scent filled the room, and she imagined Lily picking them in a field somewhere, careful and considered. A strange, pleasant sensation washed over her, and she suddenly wanted to walk around the classroom and touch all of the plants and flowers. She felt less heavy, now.

Emotion: happy

Cluster: Positive / Joy
PC1 (Valence): 3.98 Positive
PC2 (Disposition): -1.14

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.676
excitedly0.533
enthusi0.505
de0.498
๐Ÿคฉ0.495
Suppressed:
S-0.896
๐Ÿ˜ž-0.430
ness-0.410
ใใ‚Œใงใ‚‚-0.387
L-0.379