INDEX

Explanations

words related to authority and commands

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize key insights or revelations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Tokens

-0.51

lambda

-0.49

FFFF

-0.49

ãĥ¼ãĥĨ

-0.47

chuk

-0.47

]+

-0.46

emale

-0.46

 indistinguishable

-0.45

æµ

-0.45

ãĥĩãĤ£

-0.45

POSITIVE LOGITS

 Background

0.52

 largeDownload

0.50

 GOODMAN

0.47

 Dism

0.47

reb

0.46

resa

0.45

 PRESS

0.45

 DRAGON

0.45

ccording

0.44

 DARK

0.43

Activations Density 2.169%

words related to authority and commands

phrases that emphasize key insights or revelations

No Known Activations