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 chipped paint on the buildings, the cracked sidewalks, the general shabbiness of the neighborhood seemed to mirror the internal state of Elena. She was tired of feeling like a failure. She had so much potential, but seemed unable to translate any of it into success.

Sheโ€™d been avoiding eye contact with people, unable to face any kind of interaction. The yard sale was a welcome distraction. Then she saw it: the bear. Its fur was patchy and threadbare, its once-bright eyes now dull and lifeless.

A profound sadness washed over her, a deep and echoing emptiness. She stood frozen, the world around her fading into a blur. She felt like this bear, discarded and forgotten. She turned her back on the yard sale and quickened her pace, desperate to escape the painful reminder.

Emotion: worthless

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.427
S0.405
๐Ÿ˜ž0.386
ness0.339
็ผบไน0.326
Suppressed:
de-0.320
(!)-0.286
้šๅŽ-0.268
๐Ÿ˜-0.262
tc-0.255