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).

"Okay, I'm literally going to fall over," Elara groaned, her voice thick. Her back ached from hunching over her desk, the late-night conversation with "Shadow_Walker" – who, as it turned out, lived just a few blocks away – had sapped her strength.

The digital words on the screen swam before her eyes. Each breath felt like an effort. A quick meeting was proposed, just a quick coffee. A shudder ran through her at the idea of public interaction.

She leaned back in her chair, closing her eyes for a moment. The world outside, usually buzzing with life, now seemed muted, far away. The thought of facing the world felt impossibly difficult.

“This is probably a bad idea,” she mumbled, but then again, Shadow_Walker did seem interesting. Or maybe she was just being too friendly.

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
భి-0.281
เพราะ-0.277