INDEX

Explanations

numeric information such as percentages, measurements, or scores

oai_token-act-pair · gpt-3.5-turbo

key terms associated with statistics or reports

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ANS

-0.62

%:

-0.60

?????-

-0.56

Eva

-0.53

Ì¶

-0.53

geries

-0.50

iverse

-0.49

OSED

-0.48

avery

-0.48

TAIN

-0.48

POSITIVE LOGITS

of

1.58

of

1.20

 thereof

1.11

Of

1.06

OF

1.05

Of

1.01

OF

0.81

oft

0.81

ta

0.66

ofi

0.50

Activations Density 0.797%

numeric information such as percentages, measurements, or scores

key terms associated with statistics or reports

No Comments

No Known Activations