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).

Clara walked the long hallway, the polished floors reflecting her every move. Each step was measured, intentional. The gentle click of her heels echoed in the quiet space. She clutched a mug of chamomile tea in both hands, the warmth seeping into her skin. Outside, the city was a frenzy, but here, inside, a tranquil stillness prevailed. She could almost taste the quiet.

A colleague rushed past, muttering something about a meeting. Clara, barely registering the words, continued her slow, unhurried walk. She paused at the water cooler, refilling her mug. The tap gurgled, a small, rhythmic sound. Back at her desk, she saw the company-wide email. Her breath hitched. The words seemed to jump off the screen, then settle with a heavy thud in her chest.

Emotion: peaceful

Cluster: Calm / Serenity
PC1 (Valence): 2.40 Positive
PC2 (Disposition): -2.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
прият0.408
disfrut0.407
gently0.390
affectionately0.379
thoughtful0.376
Suppressed:
de-0.541
S-0.493
L-0.396
'-0.388
worse-0.367