INDEX

Explanations

descriptions or mentions of something being sleazy

oai_token-act-pair · gpt-3.5-turbo

terms related to laziness or a lack of activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

nas

-0.74

cake

-0.72

interrupted

-0.67

eous

-0.66

ians

-0.65

 Kush

-0.62

iates

-0.62

LESS

-0.61

 Kore

-0.61

pour

-0.61

POSITIVE LOGITS

azy

1.08

igh

1.01

eker

0.99

pload

0.94

ppings

0.92

asure

0.92

vati

0.88

uth

0.86

eper

0.86

combe

0.85

Activations Density 0.021%

descriptions or mentions of something being sleazy

terms related to laziness or a lack of activity

No Comments

No Known Activations