INDEX

Explanations

identification numbers and identification-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to identification and registration documents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 hindsight

-0.74

 Tend

-0.72

¥µ

-0.68

 Akin

-0.68

 Pengu

-0.68

 brink

-0.66

 layoffs

-0.65

fell

-0.64

storms

-0.64

kus

-0.64

POSITIVE LOGITS

 identifier

1.64

number

1.44

 number

1.39

 password

1.30

 identifiers

1.28

 identification

1.28

UID

1.28

Number

1.27

 address

1.27

 username

1.25

Activations Density 0.359%

identification numbers and identification-related terms

terms related to identification and registration documents

No Comments

No Known Activations