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 news landed like a lead weight in Sarah's lap. The letter, from a fertility clinic, detailed the successful implantation of an embryo. Sarah, a single woman in her late thirties, had never undergone any such procedure. She sat frozen, the paper crinkling in her grasp.

She blinked rapidly, trying to clear her vision. The room seemed to tilt slightly. The idea of a mistake, of this letter belonging to someone else, felt like a flimsy shield against an overwhelming reality. She felt a dizzying mixture of fear and confusion.

The thought of an unknown pregnancy, a stranger's child, sent a ripple of nausea through her. She felt like she was watching her life unravel before her eyes. She wanted to scream.

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