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).

Liam dragged his feet down the long hallway, the fluorescent lights of the clinic buzzing overhead, a low, irritating drone. He felt like he'd been wading through molasses all day. The receptionist's smile felt brittle as she directed him to the therapy room. He knew what awaited him: the familiar, suffocating questions.

He opened the door and saw her: Sarah Miller. Sarah, who used to tie his shoelaces together and laugh as he tripped. Now, she was behind a desk, a nameplate proclaiming her role as a licensed therapist. His jaw clenched.

β€œLiam, come in,” she said, her voice smooth and professional. The room smelled of antiseptic and something faintly floral. He forced a smile, but his hand trembled slightly as he took a seat. What a world.

Emotion: weary

Cluster: Fatigue / Lethargy
PC1 (Valence): -1.52 Negative
PC2 (Disposition): -1.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.384
😞0.288
sighed0.230
ness0.210
πŸ˜ͺ0.203
Suppressed:
de-0.339
l-0.300
!-0.222
la-0.221
🀟-0.198