INDEX

Explanations

phrases related to a series or a sequence of articles

oai_token-act-pair · gpt-3.5-turbo

references to series or installments in articles

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

animate

-0.78

enance

-0.68

$$$$

-0.67

æ©

-0.67

unts

-0.67

 wink

-0.66

 CLSID

-0.64

 Confederacy

-0.63

existent

-0.62

 disbanded

-0.61

POSITIVE LOGITS

 installments

1.02

 installment

0.99

 debunk

0.87

Interview

0.86

 series

0.80

Episode

0.79

mortem

0.79

isode

0.78

 detailing

0.76

podcast

0.76

Activations Density 0.265%

phrases related to a series or a sequence of articles

references to series or installments in articles

No Comments

No Known Activations