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

LIN

-0.66

âĿ

-0.65

Written

-0.63

 specials

-0.63

hops

-0.63

TL

-0.62

NEW

-0.60

 poisons

-0.60

Ont

-0.60

Details

-0.59

POSITIVE LOGITS

ulia

0.84

rium

0.75

rency

0.73

 Rasm

0.71

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.70

urally

0.70

uria

0.70

ardless

0.69

adium

0.68

atform

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.