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).

He knew it was coming. The rumors had swirled for months, a low hum of anxiety in the hallways. But hearing the official confirmation, seeing the cold, clinical language of the press release, still sent a jolt of panic through David. He felt the blood drain from his face, leaving him pale and shaky.

He sat on his bed, surrounded by the remnants of his life – textbooks, half-finished assignments, posters of bands he’d never see perform live again. He felt a cold dread settling in his gut, a bitter taste. He felt his hands begin to tremble.

He tried to focus on his computer screen, but the words swam before his eyes. He opened a new tab, searching for job postings, his fingers fumbling over the keyboard. The search results, a bleak landscape of uncertainty, only deepened the sense of helplessness.

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