INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ï¸ı

-0.80

Ö¼

-0.65

ges

-0.65

Ì¶

-0.62

 suspic

-0.61

 hailed

-0.60

 cheap

-0.59

 Barclays

-0.58

 Tasman

-0.57

amic

-0.57

POSITIVE LOGITS

 visitation

0.74

MpServer

0.72

geist

0.70

 Blend

0.69

Reviewed

0.69

Reviewer

0.69

volent

0.68

ocation

0.68

exit

0.68

ocations

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.