INDEX

Explanations

references to or mentions of written columns

oai_token-act-pair · gpt-3.5-turbo

references to columns or articles in a publication

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

CRIP

-0.74

à¹

-0.71

Gos

-0.70

lder

-0.67

qual

-0.65

 Romanian

-0.64

 Customs

-0.61

 Deposit

-0.61

achev

-0.60

 Thai

-0.60

POSITIVE LOGITS

 columns

1.00

 column

0.88

column

0.81

 headers

0.80

 inches

0.78

ists

0.77

oad

0.75

ftime

0.73

antry

0.73

ical

0.72

Activations Density 0.012%

references to or mentions of written columns

references to columns or articles in a publication

No Comments

No Known Activations