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 day stretched before her, an endless expanse of empty hours. Amelia wandered through her house, touching objects, as if searching for a hidden meaning she could not name. The silence pressed in, a physical thing. She felt the heavy languor that would not leave. The thought of another afternoon spent staring at the ceiling filled her with a nameless dread.

She climbed to the attic, the air growing colder as she ascended. A cold wind blew through a crack in the window frame. She surveyed the space: boxes, furniture covered in sheets, a forgotten era of lives lived.

Inside one of the boxes, she discovered a collection of vintage photographs. Smiling faces, formal poses, moments captured in time. Each photo a story, each a gateway into a different world. She felt a flicker of something new, a feeling that had been lost for a while.

Emotion: bored

Cluster: Fatigue / Lethargy
PC1 (Valence): 0.04 Positive
PC2 (Disposition): -1.11

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.387
’0.350
unremarkable0.325
दिलचस्प0.323
seemingly0.319
Suppressed:
B-0.592
P-0.304
because-0.292
因为它-0.287
ancar-0.278