© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Neuronpedia

Jacobian LensNEW

Natural Language

NEW Assistant AxisNEW Circuit TracerUPDATESteer SAE Evals Exports interp-engineNEWAPI Community Blog Privacy & Terms Contact

GPT2-SMALL · 8-TRES-DC · 1467 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
1467

INDEX

Explanations

variations of the phrase "believe it or not."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Embeds

Show PlotsShow ExplanationShow ActivationsShow Test FieldShow SteerShow Link

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enegger

-0.87

mble

-0.74

 showc

-0.68

ENCY

-0.68

 Mechdragon

-0.68

 Lumin

-0.65

theless

-0.65

lucent

-0.63

population

-0.62

 Azerb

-0.61

POSITIVE LOGITS

act

0.74

isms

0.70

isons

0.69

ope

0.68

anges

0.66

esar

0.66

hill

0.65

urn

0.65

eer

0.63

hoe

0.62

Activations Density 0.032%

No Known Activations