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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

MpServer

-0.78

HCR

-0.76

joining

-0.70

gin

-0.68

pend

-0.67

 arrog

-0.66

gr

-0.65

wings

-0.65

oj

-0.62

gon

-0.62

POSITIVE LOGITS

 Sawyer

0.67

 Recession

0.65

 Surge

0.64

 Vers

0.63

 Ended

0.61

 Hemisphere

0.61

Wol

0.60

 Wilhelm

0.59

 Weak

0.59

 Summers

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.