INDEX

Explanations

spam-related terms

oai_token-act-pair · gpt-3.5-turbo

mentions of spam in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

hani

-1.14

IST

-0.77

avery

-0.71

 Borders

-0.67

Cel

-0.65

 Templ

-0.65

 Vernon

-0.65

 Lauder

-0.63

 Syri

-0.63

 Dayton

-0.62

POSITIVE LOGITS

ming

1.39

 spam

0.95

ulent

0.85

inator

0.85

mer

0.83

ulus

0.82

bag

0.81

vertising

0.80

bags

0.79

ulence

0.79

Activations Density 0.021%

spam-related terms

mentions of spam in various contexts

No Comments

No Known Activations