INDEX

Explanations

formatting related to programming code, potentially related to strings and declarations

oai_token-act-pair · gpt-3.5-turbo

percentage values

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 exerc

-0.76

andowski

-0.73

Nau

-0.68

ITED

-0.68

ipples

-0.66

 Mellon

-0.64

Rav

-0.63

 disse

-0.63

æ©

-0.62

ãĥīãĥ©

-0.62

POSITIVE LOGITS

%%%%

1.08

AppData

0.95

reet

0.94

imate

0.83

username

0.80

lu

0.79

typ

0.79

imates

0.79

chance

0.78

%%

0.78

Activations Density 0.032%

formatting related to programming code, potentially related to strings and declarations

percentage values

No Comments

No Known Activations