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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

album

-0.64

 Herrera

-0.61

 Scand

-0.61

achable

-0.61

aqu

-0.60

icho

-0.60

ducers

-0.60

essential

-0.60

 Lovely

-0.59

urate

-0.59

POSITIVE LOGITS

ĪĴ

0.96

 veter

0.76

bilt

0.71

Â¯Â¯Â¯Â¯

0.67

alogy

0.67

 smokes

0.67

dayName

0.67

 exha

0.66

igslist

0.66

izont

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.