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

enhagen

-0.78

arus

-0.72

 ignition

-0.70

 communion

-0.68

tta

-0.65

ignant

-0.64

soc

-0.63

 ______

-0.63

icular

-0.62

oward

-0.62

POSITIVE LOGITS

ãĥ¼ãĥĨãĤ£

0.71

EG

0.68

 Rohing

0.67

Az

0.67

 Reno

0.65

 booked

0.65

Mub

0.64

milo

0.62

 Oper

0.61

 MISS

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.