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 air felt thick, heavy, as if someone had just switched off the oxygen. Her knees buckled slightly, and she grabbed onto the wrought-iron railing of the porch for support. She could feel the blood rushing in her ears, a loud, pounding rhythm. This couldn't be happening.

She moved her lips, but no words came out. The car was not there. The space was empty. She felt a strange buzzing in her head, a disorienting sensation. She looked from side to side, as if looking for something she had lost.

She closed her eyes for a moment, trying to gain her composure, but when she opened them again, the empty spot was still there, a glaring void. She felt as though she’d been plunged into a deep, confusing dream, one she couldn’t wake up from.

Emotion: dumbstruck

Cluster: Surprise / Confusion
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.71

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.326
не0.285
😶0.275
仿佛0.267
依旧0.258
Suppressed:
enthusi-0.278
la-0.264
K-0.260
wholeheartedly-0.255
쾌-0.254