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).

"Finally," Mark sighed, brushing dust from his hands. He'd been avoiding his aunt's storage unit for months. He had expected to find a jumble of family heirlooms, but instead, he felt a weight in his gut. Inside a box marked "Personal Papers" he found a letter, written in his uncle’s familiar script. The heading was devastatingly simple: “Draft – Resignation from the Police Force.”

He sat down on a folding chair, the sudden force making it creak. His uncle, the brave, upstanding officer, a man he'd looked up to his whole life...wanted to leave the force? The letter spoke of internal pressures, corruption, and the erosion of trust he had found difficult to stomach. A chill ran down his spine. The air felt thin.

Emotion: surprised

Cluster: Surprise / Confusion
PC1 (Valence): -0.58 Negative
PC2 (Disposition): -0.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.495
de0.447
la0.431
l0.323
😳0.296
Suppressed:
每天-0.269
cheerfully-0.249
ing-0.238
alleviate-0.224
gladly-0.219