INDEX

Explanations

words related to advertisements

oai_token-act-pair · gpt-3.5-turbo

occurrences and references to advertisements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 vacuum

-0.70

 chlor

-0.70

 vacant

-0.69

 Illum

-0.69

 phyl

-0.68

 worth

-0.63

 confirmed

-0.62

 Herz

-0.61

 known

-0.60

 condition

-0.60

POSITIVE LOGITS

ads

4.66

ADS

2.18

ad

2.08

adic

1.97

AD

1.56

aders

1.48

add

1.31

ady

1.28

agers

1.27

adas

1.25

Activations Density 0.010%

words related to advertisements

occurrences and references to advertisements

No Comments

No Known Activations