INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ucle

-0.75

=$

-0.74

TBD

-0.73

omial

-0.71

istics

-0.71

0000000000000000

-0.70

ãĤº

-0.68

ether

-0.67

ãĥ¼ãĥ³

-0.67

ument

-0.67

POSITIVE LOGITS

©¶æ

0.78

£ı

0.75

 exha

0.71

lear

0.69

ĪĴ

0.67

allel

0.64

RAW

0.63

 cleaners

0.63

grim

0.63

Ħ¢

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.