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

cale

-0.81

theless

-0.68

 Recon

-0.67

philis

-0.66

 Cadillac

-0.66

hang

-0.66

acebook

-0.65

Tea

-0.64

keleton

-0.64

anqu

-0.63

POSITIVE LOGITS

ovic

0.91

ovich

0.74

 posters

0.71

 stunts

0.68

 Vaugh

0.67

uv

0.65

owicz

0.65

ato

0.64

 fodder

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.