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 lawyer had just finished reading the will, and now, it was Emma's job to go through her father's belongings. His meticulous organization had always been a source of both comfort and irritation. She went through file after file, expecting the expected. Then, nestled between the insurance papers and the mortgage details, she found it: a small, spiral-bound notebook. A frisson of something she couldn't name ran through her.

The first entry read, in bold, capital letters: “CLIMB MOUNT EVEREST.” Emma choked, her hand flying to her mouth. Her father, a man who got winded walking up the stairs, wanted to climb Mount Everest? It was impossible.

She continued, the blood draining from her face. “Learn to code.” “Write a screenplay.” “Run a marathon.” Emma felt the room start to spin. The man she thought she knew… the chasm between her perception and this vibrant secret widened with each entry.

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