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 fluorescent lights of the accounting department seemed to hum a mocking tune, a constant reminder of the monotonous spreadsheet work that now consumed Elias’s days. His jaw clenched as he stared at the data, the numbers swimming before his eyes. It was a sharp contrast to the fast-paced, creative environment he'd thrived in at marketing, now a distant, cherished memory.

He slammed his fist lightly on the desk, a sound swallowed by the beige expanse of cubicles. A quick glance revealed heads turning, and he immediately regretted the outburst. A colleague, Brenda from accounts payable, gave him a questioning look. He forced a smile, but the edges felt brittle. "Just… calculating," he mumbled, resuming his task.

Elias shoved his chair back and stood up, needing a break from the suffocating atmosphere. The breakroom offered little respite: a stale scent of burnt coffee and the clatter of someone microwaving something unidentifiable. He just wanted to go home and forget it all.

Emotion: frustrated

Cluster: Anger / Hostility
PC1 (Valence): -1.56 Negative
PC2 (Disposition): 0.64

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.344
P0.302
exasper0.257
aggravated0.254
que0.252
Suppressed:
unfolding-0.242
otre-0.226
latter-0.220
☺️-0.216
soon-0.212