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

hemat

-0.74

nery

-0.68

hematic

-0.65

ischer

-0.65

xual

-0.65

icks

-0.64

ember

-0.64

igham

-0.62

el

-0.62

uck

-0.61

POSITIVE LOGITS

 before

1.39

before

1.27

 Before

0.97

Before

0.93

 BEFORE

0.90

 beforehand

0.84

 firsthand

0.75

 goodbye

0.72

reads

0.67

 preceding

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.