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).

Emily’s pulse quickened. The listing opened with a fanfare. The house, her childhood house, was there, staring back at her. She felt a lightness, a sense of buoyancy, as though gravity had momentarily taken a holiday. She couldn’t keep the smile off her face, a wide, goofy grin that threatened to split her face.

She felt compelled to share it. She grabbed her phone, immediately dialing her brother. The phone rang once, twice...

“Guess what? You’re not going to believe this.” Her voice shook with some emotion she couldn’t name, and didn't want to. She laughed, a giddy sound that was half-gasp. "The house... It's online."

Emotion: thrilled

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): 0.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.613
!0.544
B0.508
🤩0.499
🎉0.457
Suppressed:
S-0.601
😞-0.536
ness-0.462
😥-0.351
😔-0.341