© 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 · 2011 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
2011

INDEX

Explanations

terms related to natural landscapes and ecosystems

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

��

-0.77

 Extension

-0.75

 Dolphin

-0.67

driver

-0.67

arcity

-0.67

覚醒

-0.66

 VICE

-0.65

 ALWAYS

-0.65

 Width

-0.64

Dia

-0.62

POSITIVE LOGITS

 clouds

0.76

 soils

0.73

 trees

0.72

 dormant

0.72

 weeds

0.71

 gears

0.69

 roofs

0.67

 desert

0.66

 fences

0.65

 deserts

0.64

Activations Density 0.350%

No Known Activations