INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 navigate

-0.71

 Span

-0.65

 Winds

-0.64

stellar

-0.64

ictions

-0.63

 Titus

-0.62

icted

-0.61

 Canary

-0.61

inator

-0.60

zon

-0.59

POSITIVE LOGITS

guards

0.80

Letter

0.78

Secret

0.78

wcsstore

0.75

croft

0.74

Pwr

0.73

agar

0.71

gor

0.69

urg

0.67

taboola

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.