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

ences

-0.72

eric

-0.63

 approx

-0.63

icious

-0.63

ably

-0.62

 Hearth

-0.56

ric

-0.56

ington

-0.56

 Hancock

-0.56

bane

-0.55

POSITIVE LOGITS

ModLoader

0.86

displayText

0.75

 adolesc

0.72

ORE

0.70

 reluct

0.70

oing

0.70

OU

0.69

ULT

0.67

Chair

0.66

 veter

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.