INDEX

Explanations

the word "anything"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "anything."

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

irth

-0.78

pa

-0.72

ros

-0.72

grad

-0.66

ades

-0.63

inst

-0.63

asio

-0.62

bal

-0.62

ped

-0.61

ase

-0.60

POSITIVE LOGITS

 else

1.48

Else

1.37

THING

1.31

 imaginable

1.15

 Else

0.99

 remotely

0.98

 conceivable

0.98

 resembling

0.91

soever

0.91

ĪĴ

0.86

Activations Density 0.018%

the word "anything"

occurrences of the word "anything."

No Comments

No Known Activations