INDEX

Explanations

mentions of parody or advertisements

oai_token-act-pair · gpt-3.5-turbo

words or phrases related to patterns of behavior or trends in the context of media and culture

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Brill

-0.66

 Valiant

-0.65

Kob

-0.57

 stars

-0.56

shapeshifter

-0.56

 Fren

-0.55

 Hatt

-0.53

 fronts

-0.53

hower

-0.51

 corner

-0.51

POSITIVE LOGITS

enture

0.87

utical

0.82

icip

0.80

ĪĴ

0.80

opa

0.77

pmwiki

0.77

ormal

0.77

nown

0.76

aded

0.73

illance

0.73

Activations Density 0.083%

mentions of parody or advertisements

words or phrases related to patterns of behavior or trends in the context of media and culture

No Comments

No Known Activations