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

Enlarge

-0.80

Royal

-0.72

 Valiant

-0.68

Aid

-0.67

oller

-0.64

Beat

-0.63

Reviewer

-0.63

Ess

-0.63

License

-0.63

Meg

-0.62

POSITIVE LOGITS

pired

0.92

hers

0.76

well

0.74

 evidenced

0.74

phalt

0.73

 regards

0.73

ptin

0.73

eless

0.70

ĪĴ

0.69

piration

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.