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 world felt muted, as if someone had turned down the volume. His shoulders slumped, his steps echoing softly on the polished floor. The vibrant energy of the office, usually a source of motivation, felt like a jarring assault on his senses. The thought of this promotion was supposed to be a motivator.

He found himself staring at his reflection in the tinted window of his office door, barely recognizing the tired face that stared back. The prospect of the interview loomed large, a dark cloud hanging over his head. The weight of expectation felt unbearable.

"Morning, Greg!" A perky voice startled him. It was Carol from HR. He forced a smile, a gesture that felt both unnatural and exhausting. "Good morning," he mumbled, his voice a low monotone. He just wanted the day to be over.

Emotion: droopy

Cluster: Fatigue / Lethargy
PC1 (Valence): -1.15 Negative
PC2 (Disposition): -1.23

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.303
无法0.252
いつも0.238
😔0.230
:(0.228
Suppressed:
l-0.360
P-0.356
B-0.339
de-0.334
!-0.239