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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

etically

-0.75

 IMAGES

-0.74

iens

-0.73

NG

-0.72

dan

-0.67

Temp

-0.66

etics

-0.66

idel

-0.63

oxin

-0.63

imeter

-0.62

POSITIVE LOGITS

arov

0.88

Ó

0.77

 grooming

0.63

FIR

0.63

 Sessions

0.61

 streng

0.60

 Lomb

0.60

ellow

0.58

dayName

0.57

CHAT

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.