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).

Rain hammered against the studio window, mirroring the storm brewing inside Maya. She'd been staring at her half-finished canvas for an hour, the vibrant colours seeming to mock her inertia. She picked up a brush, hesitated, then let it fall back onto the easel with a dull thud. Her head felt heavy, each thought a lead weight.

A message popped up on her phone. A photo: a man’s back, adorned with a near-exact replica of her “Fractured Cityscape” design. She could barely breathe. The air in the room seemed thick, suffocating. She felt a profound weariness, a bone-deep hollowness.

She wanted to scream. Instead, she let out a sigh, the sound lost in the drumming of the rain.

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