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 air was thick with humidity, clinging to Emily like a second skin. Her steps were slow, her gaze vacant, as she wandered through a forgotten corner of her backyard. The sun, a hazy disc behind the clouds, offered no warmth. She was swamped in the day's heaviness.

She tripped over something buried in the soft earth. She paused, then, with a long sigh, started digging. She unearthed a small, tin box. She had no recollection of it. Or its purpose. She opened it and examined the contents.

Inside were a couple of colorful glass marbles and a faded copy of a childhood book. The marbles and book brought forth no recognition, no spark of nostalgia. She closed the box and looked at the ground. She was incredibly tired.

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