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 information arrived at the worst possible time, during a particularly tough final exam. David glanced at his phone, the email notification flashing in the corner. His heart sank. He could barely focus on the questions in front of him.

He’d put everything on hold for this degree. His social life, relationships, even his health. The promise of a better future had kept him going. Now, that future was gone. The ink on the exam paper blurred.

He felt a profound sense of failure, not just for himself but for his family, who had supported him. His carefully constructed facade of optimism crumbled. The exam room, once a symbol of his aspirations, felt like a prison. He knew he'd fail the test.

Emotion: dispirited

Cluster: Sadness / Despair
PC1 (Valence): -1.88 Negative
PC2 (Disposition): -0.98

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.398
S0.348
’0.325
ness0.320
无法0.308
Suppressed:
de-0.297
l-0.256
πŸ‘-0.251
sonra-0.240
aud-0.240