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, though muted by the grimy window, felt like a spotlight on her face. Every sense seemed heightened, every sound amplified. She found herself picking at a loose thread on her coat, a nervous habit she hadn't realized she still had. The shop was a blur, a collection of forgotten things.

She stood at the counter, her shoulders slouched, the effort of remaining upright almost overwhelming. The shop owner, a man whose face was a roadmap of disappointments, leaned over the counter. He said something, but the words seemed to float away.

Then, it caught her eye. In the shadows, a glint. A familiar shape. She walked towards it, her legs feeling like lead. Her grandmother's ring, a tiny beacon of memory in a sea of forgotten treasures. She just stood there for a long moment, simply staring.

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
à°­à°¿-0.214