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

isance

-0.77

atives

-0.74

 collectors

-0.70

 bargain

-0.68

 auction

-0.65

['

-0.65

found

-0.64

 coupons

-0.64

 proc

-0.64

 auctions

-0.63

POSITIVE LOGITS

 Liang

0.76

ividual

0.75

Disclaimer

0.74

 Qiao

0.73

 Deaths

0.71

 IMAGES

0.71

ral

0.71

methyl

0.70

 Awoken

0.67

NAME

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.