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 cheerful chatter of the coffee shop grated on Isabella's nerves. She stared at the email on her phone, the words a cold slap. The scholarship, the key to her medical school ambitions, was now…gone. A clerical error. Not hers. The blood drained from her face.

She felt as if the ground had opened up beneath her. Her carefully constructed future, reduced to a single sentence on a screen. She struggled to take a breath, her chest tightening with each passing moment.

She gripped her phone so tightly that her knuckles turned white. She felt a burning in her eyes, a stinging sensation she fought to suppress. She couldn't cry. Not now.

She needed to think. She needed a plan. She rose and practically ran to the bathroom, desperate to escape the judgement of the public eye.

Emotion: panicked

Cluster: Fear / Anxiety
PC1 (Valence): -2.44 Negative
PC2 (Disposition): -0.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.317
😫0.294
😖0.285
😰0.285
escape0.283
Suppressed:
la-0.432
a-0.374
de-0.296
楽し-0.262
happy-0.258