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 newspaper was open on the breakfast table, the headline screaming at her. "Clara Bellweather: The Woman Who Wouldn't Conform." It was *her* face staring back from the photograph, a picture she didn't recall taking. She felt a knot tighten in her chest.

She reached for her tea, her hand shaking so badly that she sloshed it. The article went into detail about her various eccentricities: her love of collecting buttons, her refusal to own a cell phone, her tendency to speak in rhymes. She hadn't told anyone these things, except, perhaps, her therapist.

She felt a sudden urge to flee. She grabbed her coat, not even bothering to finish her breakfast, and stumbled out the door. The world suddenly felt tilted, uncertain.

Emotion: puzzled

Cluster: Surprise / Confusion
PC1 (Valence): -1.17 Negative
PC2 (Disposition): -0.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.453
🤔0.421
🤔0.402
🧐0.328
a0.294
Suppressed:
ecstatic-0.265
gladly-0.238
swiftly-0.237
alleviate-0.234
wholeheartedly-0.232