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).

"Did you see the new drawing?" Amelia's voice sounded strained. She avoided making direct eye contact with her husband, Mark. He looked at the drawing, a portrait of a smiling girl alongside a figure with too many limbs and eyes that were too large. The creature he’d named ‘Blink’ when he was six.

He felt his heart start to race. A wave of heat washed over him, followed by a cold sweat. He remembered the feeling of being alone.

He couldn't meet Amelia's gaze. He found himself picking at a loose thread on the sofa. He didn’t want to talk about it. He didn't want to remember.

Later, he claimed he was going for a walk, leaving his wife to wonder when he would return.

Emotion: unsettled

Cluster: Fear / Anxiety
PC1 (Valence): -2.33 Negative
PC2 (Disposition): -0.65

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.382
😞0.309
ness0.301
😣0.256
😔0.248
Suppressed:
de-0.721
la-0.597
a-0.365
"-0.299
l-0.274