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

ãĥ£

-0.76

ANY

-0.66

itaire

-0.65

 diarr

-0.64

IPM

-0.63

 skelet

-0.61

eteria

-0.61

raw

-0.60

annels

-0.60

 Antar

-0.59

POSITIVE LOGITS

Downloadha

0.86

leased

0.75

Rated

0.72

rar

0.71

lander

0.71

iership

0.71

bilt

0.69

phabet

0.67

wreck

0.67

woman

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.