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).

"Could you, please, be mindful of the noise?" Clara stammered through the closed door. The response was a raucous burst of laughter and music that seemed to mock her. She rubbed her arm, a gesture she couldn't explain.

She returned to her small apartment. The sound of her neighbour's party permeated everything. She walked from room to room, searching for a place of peace, but there was nowhere. Each corner offered nothing but echoes.

She closed her eyes, trying to block out the noise. Her chest tightened, and she felt a deep, aching loneliness. The party continued, and she was adrift, marooned in the silence of her own life, in contrast.

Emotion: trapped

Cluster: Sadness / Despair
PC1 (Valence): -3.02 Negative
PC2 (Disposition): -0.80

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.573
ness0.401
无法0.366
hopeless0.360
😞0.353
Suppressed:
la-0.740
de-0.376
l-0.339
several-0.323
excitedly-0.322