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

Row

-0.80

ink

-0.79

SPONSORED

-0.78

olkien

-0.73

ologically

-0.72

species

-0.72

olia

-0.71

DragonMagazine

-0.70

ump

-0.70

elt

-0.69

POSITIVE LOGITS

 Corpus

0.78

sur

0.72

 relay

0.69

 pier

0.68

 Portsmouth

0.68

Ern

0.67

 Sabbath

0.66

 Arsenal

0.66

 Overse

0.66

 Shore

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.