© 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 · 2882 ｜ Neuronpedia

Home
GPT2-Small
Transcoders Residuals
8-TRES-DC
2882

INDEX

Explanations

terms related to recruitment and selection processes

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

rodu

-0.81

 promot

-0.79

ufact

-0.72

 rehabilit

-0.68

iversal

-0.68

 replication

-0.67

 subsequ

-0.64

 Facilities

-0.64

erial

-0.64

 vocational

-0.64

POSITIVE LOGITS

��

0.70

 smelled

0.70

 Photograph

0.69

Tips

0.69

PHOTOS

0.65

 Oops

0.64

ATES

0.64

 Venice

0.63

�

0.63

oops

0.62

Activations Density 0.243%

No Known Activations