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).

Her feet ached, her shoulders were stiff, and her brain felt like cotton. The office had been a pressure cooker, the commute a grueling ordeal. All she wanted was to collapse into her bed and erase the day.

But then, a notification. It was from Ben, her online companion. He always knew how to cheer her up, with his witty remarks and his endless supply of amusing stories. He was her escape, a source of lightness in a world that often felt suffocating.

He mentioned that he was taking his dog to the park. The park he frequented. The park she lived next to. She messaged him back, a playful quip about the local squirrel population. He responded immediately. A sudden, unexpected realization hit her like a lightning bolt: they were neighbors. The thought, suddenly, was invigorating.

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