INDEX

Explanations

words or phrases related to advertisements

oai_token-act-pair · gpt-3.5-turbo

occurrences of advertisements

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

bred

-0.77

joy

-0.73

Kut

-0.71

 Myster

-0.66

 invite

-0.66

pled

-0.64

 motivating

-0.64

 Ultr

-0.64

ker

-0.63

 Koch

-0.63

POSITIVE LOGITS

ADVERTISEMENT

1.34

iciary

1.06

vertising

0.91

interstitial

0.85

ħĭ

0.84

swer

0.82

flush

0.81

 citiz

0.81

taboola

0.81

Skip

0.80

Activations Density 0.004%

words or phrases related to advertisements

occurrences of advertisements

No Comments

No Known Activations