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 first drill of the morning sliced through the silence like a hot knife through butter. Elias, normally a creature of habit and routine, found himself grinning. He threw off his covers, the sudden movement causing his chest to swell. He leaped out of bed, adrenaline pumping. He needed to get moving. He threw on his oldest jeans and a faded band t-shirt, ignoring the time on the clock. He hadn't felt this alive in months.

He swung open the front door, the cool morning air washing over him. The sounds next door, previously an annoyance, now sounded like a symphony. He inhaled deeply, a manic energy course through him. He bounded down his driveway, a spring in his step, and stood at the fence, watching the demolition. He clapped his hands together, an impulse he couldn’t suppress.

“Morning!” he boomed to the construction crew, his voice louder than he intended.

Emotion: invigorated

Cluster: Positive / Joy
PC1 (Valence): 3.28 Positive
PC2 (Disposition): 0.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.452
(!)0.439
🎉0.434
invigor0.416
enthusi0.395
Suppressed:
😞-0.524
😔-0.428
😢-0.404
😥-0.393
que-0.391