INDEX

Explanations

mentions of a specific username or phrase probably related to online content

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "fe."

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

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

 gorilla

-0.68

 destiny

-0.64

 tray

-0.64

 Akron

-0.64

SAS

-0.63

 trash

-0.62

 outnumbered

-0.62

 insert

-0.62

UTC

-0.61

ATL

-0.60

POSITIVE LOGITS

fe

4.43

FE

1.63

Fe

1.48

fle

1.37

fing

1.36

fu

1.33

ffe

1.26

fen

1.23

1.20

fed

1.19

Activations Density 0.013%

mentions of a specific username or phrase probably related to online content

occurrences of the word "fe."

No Comments

No Known Activations