INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

actionDate

-0.83

mega

-0.79

oteric

-0.76

 guiActiveUn

-0.67

actor

-0.67

FactoryReloaded

-0.66

venture

-0.66

soType

-0.66

ilage

-0.66

ung

-0.65

POSITIVE LOGITS

 minimized

0.66

accompan

0.65

 Mean

0.61

 Horton

0.61

added

0.59

sticks

0.59

 Woodward

0.59

 Becker

0.58

 indicators

0.57

 Eleven

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.