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

Negative Logits

igger

-0.73

 ignition

-0.67

 proofs

-0.66

 noses

-0.65

iably

-0.64

ancial

-0.64

incoln

-0.63

stown

-0.63

 merger

-0.62

sect

-0.62

POSITIVE LOGITS

é¾įå¥ĳå£«

0.85

DragonMagazine

0.85

rb

0.82

ãĥ¼ãĤ¯

0.71

phies

0.71

ä»£

0.70

åĮ

0.70

saf

0.69

CVE

0.67

ãĤ©

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.