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 swam before Amelia's eyes. The vibrant chaos of the art gallery, usually a source of inspiration, now felt overwhelming, a kaleidoscope of meaningless shapes and colors. Across the room, Leo, the celebrated "art collector" and her long-term partner, was explaining he worked as a bank teller. Her breath hitched in her throat, a sharp, painful intake. Her fingers twitched uncontrollably.

Her ears roared, the sound amplifying every minor noise into a cacophony. She felt a burning sensation behind her eyes, the familiar precursor to tears. She wanted to scream, to shatter the placid facade that Leo had carefully constructed, but her voice was trapped. The gallery's bright lights seemed to throb, mirroring the frantic pulse in her temples. She took a step back, bumping into a sculpture, nearly knocking it over.

Emotion: hysterical

Cluster: Fear / Anxiety
PC1 (Valence): -3.19 Negative
PC2 (Disposition): -0.45

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.401
่ƒ€0.401
์ ์ 0.396
๐Ÿ˜ซ0.377
ใใ‚Œใงใ‚‚0.373
Suppressed:
la-0.672
a-0.618
de-0.541
๐Ÿ˜Ž-0.370
๐Ÿ˜Ž-0.359