INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

illegal

-0.76

aucas

-0.74

Lt

-0.69

itars

-0.67

 Algeria

-0.66

uld

-0.65

 Angola

-0.64

astern

-0.64

eties

-0.63

aft

-0.63

POSITIVE LOGITS

 tremend

0.80

KEY

0.68

ONSORED

0.67

 LEVEL

0.66

iard

0.66

LAB

0.65

kered

0.64

ideshow

0.64

ORY

0.63

 pause

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.