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).

Rain lashed against the windows of Leo's apartment, mirroring the turmoil in his stomach. He hadn't touched his breakfast cereal. The news about Maya... it just sat with him, a heavy weight. He’d driven past her “office” that afternoon, and there, in place of the gleaming glass facade of a marketing firm, was a dingy, boarded-up storefront. He felt a profound sense of disappointment, a dull ache in his chest.

He stared at the cereal, the milk beginning to separate. He should call her. He *needed* to call her. But the phone felt like a lead weight in his hand. He’d loved hearing her stories about the high-pressure meetings, the creative brainstorming, the power lunches. Now, they felt hollow, replaced by a strange, unsettling emptiness.

He finally pushed the bowl away, the clatter echoing in the silent apartment. He needed air, needed to be anywhere but here.

Emotion: gloomy

Cluster: Sadness / Despair
PC1 (Valence): -1.68 Negative
PC2 (Disposition): -1.37

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.368
无法0.319
😔0.300
いつも0.270
쓸0.255
Suppressed:
la-0.432
l-0.419
de-0.398
B-0.294
!-0.259