© 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 · 1492 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
1492

INDEX

Explanations

quoted speech or dialogue

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

 Tanz

-0.74

 Manz

-0.69

 Diversity

-0.69

 Barr

-0.67

 Reload

-0.66

 Tier

-0.66

 Minor

-0.65

 magnitude

-0.65

dB

-0.64

 Battles

-0.63

POSITIVE LOGITS

too

1.08

highly

1.08

pretty

1.08

puff

1.07

inspired

1.06

gar

1.06

absolutely

1.05

fed

1.04

picking

1.04

bur

1.04

Activations Density 0.041%

No Known Activations