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 air hung heavy and still, thick with the scent of old paper and forgotten things. Maria felt as if she were moving through molasses, each movement a struggle. Her mind felt foggy, her concentration shot. The day had taken its toll; she was aware that she needed to be still.

She drifted, her gaze unfocused, until a flash of familiar brown caught her eye. The bear. Her bear, Cosmo. She hadn't seen him in over twenty years. A wave of melancholy washed over her. She knew she needed to sit down, to be quiet, to do nothing.

She reached for him, her hand shaking slightly. The seller, a woman with a mountain of tangled jewelry, rattled off a price. Maria barely registered it. She closed her eyes, clutching Cosmo tightly.

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