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 in the training room felt thick, almost suffocating. He watched as the new recruit, Ben, took notes, his expression a mixture of concentration and enthusiasm. He was supposed to be imparting his wisdom, sharing his expertise. But each sentence felt like a betrayal. His own past played on a loop.

He took a slow sip of water, trying to swallow the lump in his throat. He felt the sting of his actions, and the weight was heavy. He knew he had made a mistake, that he had chosen the wrong path.

The evening was a blur of shadows. The familiar sights and sounds of the city offered no solace. The feeling was a constant, a dull ache that followed him everywhere. He thought of his family, the people he was supposed to protect, and the weight of their expectations.

Emotion: remorseful

Cluster: Shame / Guilt
PC1 (Valence): -1.57 Negative
PC2 (Disposition): 1.09

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.656
L0.580
悔0.463
remorse0.416
been0.398
Suppressed:
de-0.444
(!)-0.309
имеется-0.289
oldukça-0.287
😍-0.285