© 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 · 2498 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
2498

INDEX

Explanations

references to titles of works or notable names

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

¯¯¯¯¯¯¯¯

-0.66

ItemThumbnailImage

-0.65

 rods

-0.64

fax

-0.63

adal

-0.62

liga

-0.62

ب

-0.61

anooga

-0.61

berries

-0.61

★

-0.59

POSITIVE LOGITS

 Meet

0.67

 POLITICO

0.65

 CNBC

0.64

edient

0.63

 VICE

0.62

Vs

0.62

ESV

0.61

EW

0.60

 Exclusive

0.60

 Entertainment

0.58

Activations Density 2.110%

No Known Activations