© 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 · 673 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
673

INDEX

Explanations

negative assertions or denials

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

ebook

-0.76

ermott

-0.75

aido

-0.71

renheit

-0.68

rack

-0.67

)</

-0.67

monds

-0.65

inav

-0.65

ebin

-0.63

EStream

-0.61

POSITIVE LOGITS

 alive

0.68

 unexplained

0.65

bip

0.63

 issues

0.63

issues

0.62

witch

0.60

Joined

0.58

 parity

0.57

 conducive

0.57

 battling

0.56

Activations Density 0.107%

No Known Activations