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).

“Ugh,” she muttered, staring at her reflection. Ms. Baker felt a profound lack of get-up-and-go. The world felt muffled, as if filtered through cotton. The thought of dealing with noisy teenagers felt like a herculean task. The silence of her empty home felt welcoming compared to the prospect of her hectic workday.

A former student, Liam, whom she had taught a few years prior, had sent a package. A massive, elaborately constructed puzzle lay before her. A note: "For when you need something to focus on."

She sat down, opening the box. The intricate pieces seemed to beckon, promising a temporary escape. Her fingers moved, connecting one piece to another, and she slowly felt herself gaining some mental clarity.

Emotion: sluggish

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.56 Negative
PC2 (Disposition): -1.03

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
sluggish0.263
dreaded0.258
似乎0.257
буд0.256
barely0.254
Suppressed:
P-0.384
B-0.379
భి-0.246
羞-0.243
ຮ-0.238