INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

xual

-0.83

 Bezos

-0.69

Adult

-0.68

lord

-0.65

YING

-0.65

 Spoon

-0.65

 Alchemist

-0.62

 Valiant

-0.61

stocks

-0.60

 attachments

-0.59

POSITIVE LOGITS

imar

0.75

 inaug

0.69

idates

0.67

emer

0.66

iani

0.65

 chall

0.64

perse

0.64

eworks

0.62

 Manhattan

0.62

eper

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.