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).

Inside the library, the air was thick with the scent of old paper and the hushed whispers of students. Emily, hunched over a textbook, felt the familiar ache in her lower back. The late night study sessions had taken their toll, and a deep languor had settled over her, making it difficult to concentrate. She felt detached, as if viewing the world through a frosted pane.

Suddenly, a tap on her shoulder. Her friend, pointing at a laptop screen. The words on the screen swam before her eyes, and she barely registered the email: "Your essay is being featured!" It was her own essay. Her mind struggled to comprehend. A sense of inertia washed over her, making her movements feel sluggish and deliberate.

Emotion: sleepy

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.86 Negative
PC2 (Disposition): -1.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😴0.318
weary0.310
fatigue0.296
💤0.283
似乎0.280
Suppressed:
B-0.343
C-0.294
because-0.293
à°­à°¿-0.281
เพราะ-0.277