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

ABE

-0.79

 defe

-0.75

Ens

-0.69

 Liberties

-0.67

 Protect

-0.67

²¾

-0.65

 Ples

-0.65

 Flag

-0.62

yip

-0.62

Bye

-0.62

POSITIVE LOGITS

ums

0.93

uum

0.90

usha

0.77

hea

0.76

heit

0.76

zes

0.75

ractor

0.73

rolled

0.70

packs

0.70

IUM

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.