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).

Maya traced the looping signature at the bottom of the letter with a trembling fingertip. The penmanship was instantly recognizable, her mother's. The letter, typed on stationary with the logo of a hospital, was a declaration of leaving her job as a nurse. "Insufficient resources to ensure proper care" was the reason cited.

She had found it tucked inside a photo album, among pictures of her as a toddler. Everything she had known about her mother’s career – the pride she took in her profession, the late nights, the unwavering dedication – it now felt skewed. How long had she felt this way? What had she left behind?

She began to pace her kitchen, the letter a crumpled mass in her hand. The sunlight streaming through the window seemed too bright, the sounds of the city outside too loud. It all felt… false, somehow.

Emotion: perplexed

Cluster: Surprise / Confusion
PC1 (Valence): -1.42 Negative
PC2 (Disposition): -0.16

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🤔0.330
L0.300
🤔0.296
S0.273
que0.264
Suppressed:
மகி-0.227
rest-0.222
ecstatic-0.212
alleviate-0.210
uplifting-0.196