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).

He walked in slowly, shoulders hunched, the weight of the world pressing down. The door’s bell jingled a mournful tune. The shop seemed to shrink around him, the walls closing in. He barely registered the indifferent glance of the shop owner, who was busy polishing a revolver. He just needed this done.

He circled the room, his gaze skimming over the rows of forgotten objects. Everything seemed a meaningless jumble, a testament to the passage of time. He was close to just leaving, giving up.

Then, his eyes caught on something. A familiar sparkle in a display case. A delicate gold band with a single, flawless stone. His grandmother’s ring. He recognized it instantly. He felt a sudden, sharp pang of something he couldn’t name. A strange mix of relief and something else, something deeper. He reached for it, his hand trembling.

Emotion: tired

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.92 Negative
PC2 (Disposition): -1.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
weary0.254
fatigue0.249
😴0.238
sighed0.237
barely0.230
Suppressed:
C-0.268
because-0.225
B-0.219
🤟-0.217
భి-0.214