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

ibaba

-0.81

quickShipAvailable

-0.76

forts

-0.71

nings

-0.71

claimer

-0.71

taboola

-0.70

venth

-0.70

 coasts

-0.69

insured

-0.69

inventoryQuantity

-0.69

POSITIVE LOGITS

 Ridley

0.91

reat

0.69

 Beet

0.64

 Mane

0.63

 Mechdragon

0.62

 Guth

0.62

 Patterson

0.62

 accompl

0.61

FUL

0.60

ALP

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.