INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

hog

-0.76

.","

-0.71

 hassle

-0.70

ribute

-0.69

ratom

-0.65

orea

-0.65

 hatch

-0.65

 cleaner

-0.64

 input

-0.64

eree

-0.63

POSITIVE LOGITS

Downloadha

0.79

 Occupations

0.78

 Colleges

0.74

Footnote

0.74

 Presumably

0.73

 BART

0.72

 Arkham

0.71

 Though

0.71

 Likewise

0.70

Yog

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.