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 vibrant splashes of paint that had filled the walls of Emily's studio, a space that was once a source of creative energy, now felt like a constant reminder. The once-bright palette of colors seemed dull. Her hand trembled as she tried to brush the canvas.

Her daughter, Ruby, painted with a fearless abandon, her art a kaleidoscope of color and emotion. Ruby's work hung proudly in galleries. Emily, by contrast, hadn’t painted in months, her brushes gathering dust.

"It's finished!" Ruby announced, her voice filled with pride. Emily glanced over, unable to meet her daughter's confident gaze. The feeling was a deep ache. She forced a smile and said, "It's beautiful."

Emotion: stuck

Cluster: Sadness / Despair
PC1 (Valence): -2.17 Negative
PC2 (Disposition): -0.72

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.487
缺乏0.401
L0.361
每天0.348
stagn0.348
Suppressed:
l-0.472
la-0.346
随后-0.322
undamaged-0.268
C-0.263