© 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 · 2654 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
2654

INDEX

Explanations

terms related to security organizations and their roles

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

ña

-0.65

iddler

-0.63

arian

-0.63

omb

-0.61

edIn

-0.60

uese

-0.58

obyl

-0.58

-0.57

ⓘ

-0.57

erous

-0.56

POSITIVE LOGITS

PLIED

0.77

 Clancy

0.77

grain

0.68

estern

0.66

 agencies

0.64

EA

0.64

clear

0.64

click

0.64

 bets

0.63

riages

0.62

Activations Density 0.014%

No Known Activations