INDEX

Explanations

official Twitter accounts or mentions of Twitter interactions

oai_token-act-pair · gpt-3.5-turbo

references to social media accounts, particularly Twitter accounts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NetMessage

-0.83

 Rounds

-0.67

GGGGGGGG

-0.67

ced

-0.65

instein

-0.64

GGGG

-0.63

advertising

-0.63

xon

-0.62

ugh

-0.62

ithing

-0.61

POSITIVE LOGITS

ancy

0.93

lain

0.91

 balances

0.88

 username

0.88

 handles

0.84

 accounts

0.83

 holder

0.83

 account

0.81

"@

0.79

 passwords

0.76

Activations Density 0.021%

official Twitter accounts or mentions of Twitter interactions

references to social media accounts, particularly Twitter accounts

No Comments

No Known Activations