INDEX

Explanations

names or terms with unusual characters

oai_token-act-pair · gpt-3.5-turbo

complex or technical terminologies and phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ealous

-0.56

smoking

-0.53

eu

-0.49

udeb

-0.48

upper

-0.46

ãĤ¨ãĥ«

-0.45

elvet

-0.45

say

-0.45

Supplement

-0.43

apo

-0.43

POSITIVE LOGITS

pse

0.51

estyles

0.48

hens

0.44

 Morty

0.44

ptin

0.43

 artif

0.43

 mathemat

0.42

 Jinn

0.42

 fixme

0.42

 autom

0.42

Activations Density 7.543%

names or terms with unusual characters

complex or technical terminologies and phrases

No Comments

No Known Activations