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 had lost its colour. Everything seemed washed out, bland. The incessant doorbell, a jarring intrusion into the monotone existence. Leo stumbled to the door. Standing on the porch was Mr. Blake, his face a mask of false cheer.

"Leo, good morning!" Mr. Blake chirped. "I wanted to talk about that old pine. The needles are a bit of a nuisance, and it's getting a little close to my property." He gestured toward the tree, his smile widening slightly.

Leo felt a deep ache in his chest, a sense of apathy that seemed to permeate his very being. The thought of any form of engagement, of making a decision or taking any action at all, felt utterly overwhelming. He just stared at the tree, the image a blur, and wished he could just drift away. The weight of the moment was almost unbearable.

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