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).

He hadn't slept well since seeing his essay. He sat at his kitchen table, staring at the printout. He’d envisioned a future, an opportunity, a beginning. This felt like the end.

His hands shook as he reached for the coffee cup. He felt a wave of nausea. He felt betrayed by his own words, each sentence now a shard of glass.

He knew he should be proud. But all he felt was this crushing sensation. The weight of exposure. He felt a desire to obliterate it all, to erase the words, to disappear. He tossed the paper down, and his tea spilled and spread, leaving a dark stain that matched the one spreading across his chest.

Emotion: terrified

Cluster: Fear / Anxiety
PC1 (Valence): -2.97 Negative
PC2 (Disposition): -0.62

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.357
worse0.350
无法0.334
😰0.333
sickening0.322
Suppressed:
la-0.676
a-0.392
de-0.371
"-0.344
happy-0.309