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).

Eliza nearly dropped the mug of tea she held. Her jaw went slack, and she had to grip the counter for support. The drawing, a crude but vibrant depiction of a grinning, one-eyed monster, was undeniably Barnaby. Barnaby, her childhood companion, the boogeyman under her bed. The memory of him, a comforting evil, flooded back.

"Who's that, sweetie?" she managed, her voice a strained whisper. Her little nephew, Timmy, beamed up at her.

"Bartholomew!" he declared, his voice filled with glee. He then added, "He says he misses you." A frisson of something she couldn't name ran down Eliza's spine. It wasn't fear, exactly. It was… something else, something she hadn't felt in a very long time.

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