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 city noise grated on Emily’s nerves. The bright lights, the busy streets – it was all too much. She pulled her scarf tighter, as if to shield herself from the sheer intensity of the world. She walked through a haze, as if everything was too bright and she was squinting.

It was time to get some help. The new therapist had a good reputation.

She walked in and saw it. The man she once knew as Kevin. Kevin, who once made her life a misery, was now “Dr. Kevin O’Connell, PhD.” A wave of nausea washed over her.

"Emily, take a seat," Kevin said, his gaze steady. She felt her face burn.

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