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

phabet

-0.79

rification

-0.71

atro

-0.70

entin

-0.70

 Liberties

-0.69

 mosqu

-0.68

rush

-0.67

 princ

-0.66

ragon

-0.64

 Franks

-0.62

POSITIVE LOGITS

soType

0.82

DVD

0.69

soDeliveryDate

0.69

ADA

0.67

PokÃ©

0.67

ById

0.66

Ì¶

0.65

catentry

0.65

>>>

0.64

 Elias

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.