INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

administ

-0.77

iasis

-0.71

cise

-0.71

 externalToEVAOnly

-0.68

govern

-0.68

},{"

-0.66

iquid

-0.65

osed

-0.64

ooth

-0.63

cens

-0.63

POSITIVE LOGITS

eln

0.90

0.78

 daring

0.67

quit

0.66

Psy

0.63

Gat

0.62

ModLoader

0.62

 Dynam

0.61

oire

0.61

 quir

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.