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 restaurant was almost empty. Chef Yumi usually thrived on the buzz of a full dining room. Now, she felt exposed, vulnerable. The review – a cruel dissection of her innovative tasting menu – seemed to hang in the air, a physical weight.

She found herself meticulously cleaning the already spotless counter, scrubbing with unnecessary force. Her hands ached, her knuckles white. The clatter of a single fork being dropped sent a jolt of panic through her.

The waiter, David, approached, his expression cautious. "Chef, are you…okay?" Yumi took a deep breath, fighting back the unwelcome sting in her eyes. "Fine," she managed, her voice barely above a whisper.

Emotion: sensitive

Cluster: Fear / Anxiety
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.75

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
ness0.304
😞0.282
😥0.272
😣0.271
own0.264
Suppressed:
la-0.679
de-0.660
a-0.281
"-0.280
(!)-0.242