© 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 · 1734 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
1734

INDEX

Explanations

numeric values associated with large quantities of people

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

onde

-0.66

chy

-0.62

 arcs

-0.60

ARC

-0.60

 VIDEOS

-0.59

hetti

-0.58

andon

-0.57

 tiss

-0.57

anu

-0.57

aired

-0.56

POSITIVE LOGITS

clud

0.76

lude

0.63

abase

0.63

probably

0.62

ady

0.61

alyst

0.61

ector

0.58

0.58

eston

0.57

anooga

0.57

Activations Density 0.026%

No Known Activations