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 dusty attic was the only place Marcus felt he could breathe, or at least, be left alone. He ran his hand over the cobweb-draped furniture, barely registering its presence. Sunlight, fractured by the attic window, cast long shadows. A small box, tucked away in a corner, contained a stack of letters.

They were from his father, addressed to a woman named Emily, and a child named Clara. His heart felt as though it had been lead-filled. He sunk to the floor, the letters scattered around him like fallen leaves. A half-sister. A family he didn't know existed. The joy of discovery was not present in him.

He felt utterly unmoored, adrift in a sea of unspoken truths and buried secrets. He rested his head against the cold, wooden floor.

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