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 digital glow of the university website illuminated Sarah's face as she scrolled through the example essays. There it was: hers. The essay she'd poured her heart into, the one about her struggles and triumphs. The corners of her mouth slowly curled upwards.

Her phone buzzed, a text from her friend asking if she wanted to go for a run. “Yes!” she texted back, immediately grabbing her running shoes. The cool air hit her face, and with it, a burst of energy. Her legs flew.

The world seemed to sharpen as she ran, the colors more vivid, the sounds clearer. She didn't feel tired at all.

Emotion: inspired

Cluster: Positive / Joy
PC1 (Valence): 2.98 Positive
PC2 (Disposition): 0.04

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.563
invigor0.465
newfound0.446
هذ0.414
используя0.406
Suppressed:
😞-0.481
不-0.431
worse-0.428
😣-0.400
fact-0.389