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

awed

-0.74

 Zhou

-0.67

rock

-0.64

ileaks

-0.61

Md

-0.59

aging

-0.59

 Closing

-0.59

rior

-0.58

asso

-0.57

 diapers

-0.57

POSITIVE LOGITS

caster

0.78

uala

0.78

iferation

0.76

*/(

0.75

REDACTED

0.74

ãĤ¼ãĤ¦ãĤ¹

0.73

pmwiki

0.70

Interstitial

0.67

elligent

0.66

ãĥĦ

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.