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 scent of freshly cut grass mingled with the aroma of barbecue as I leaned against the fence, a cold beer in my hand. My neighbor, Mr. Henderson, approached, his eyes twinkling behind his glasses. He wanted to cultivate a small herb garden, a request that came across as endearing, rather than intrusive.

"Morning, Tom!" he chirped, already smiling. "Got a minute?" He gestured towards the patch of unused land near the back fence.

A slow smile crept across my face. I took a deliberate sip of beer. My shoulders relaxed. He explained the specifics – how he wanted to grow basil, mint, and rosemary.

"Absolutely, George," I said, my voice mellow. "Just point me to the area you're thinking of, and we'll get it sorted."

Emotion: patient

Cluster: Passivity
PC1 (Valence): 2.71 Positive
PC2 (Disposition): 0.40

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
rest0.467
eventually0.415
leisurely0.369
慢慢0.357
a0.355
Suppressed:
无法-0.432
even-0.428
인해-0.421
無法-0.415
😣-0.391