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).

Dust motes danced in the late afternoon sun as Amelia wrestled the shovel into the stubbornly compacted earth. A glint of metal caught her eye. *Great,* she thought, wiping sweat from her brow. More digging. She'd been looking for a lost sprinkler head, not some buried treasure. Her back ached, a low thrum of frustration building in her chest.

She scraped away the remaining dirt, revealing a tarnished metal box. The lid was held fast by rust. With a grunt, she pried it open, the hinges screaming in protest. Inside, nestled amongst yellowed newspaper, was a collection of bizarre objects: a cassette tape, a lava lamp, and a sealed letter addressed to… herself. *Seriously?*

Emotion: annoyed

Cluster: Anger / Hostility
PC1 (Valence): -0.59 Negative
PC2 (Disposition): 1.27

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.376
😠0.251
骂0.249
🤬0.245
fucking0.239
Suppressed:
own-0.352
渗透-0.251
gradually-0.234
unfolding-0.226
超越-0.224