INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

peak

-0.96

Publisher

-0.82

gow

-0.80

Leader

-0.79

Born

-0.78

jac

-0.75

spe

-0.75

hovah

-0.75

 VIDEOS

-0.72

Kay

-0.71

POSITIVE LOGITS

 unse

0.65

 damn

0.65

 heck

0.65

oot

0.64

 Tigers

0.64

iotic

0.63

 contag

0.63

 blot

0.62

 flux

0.62

 damned

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.