INDEX

Explanations

privacy-related terms

oai_token-act-pair · gpt-3.5-turbo

references to privacy and related policies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

xual

-0.92

kell

-0.72

cki

-0.71

annis

-0.70

Production

-0.67

×Ļ×

-0.66

flat

-0.66

ings

-0.65

enegger

-0.64

urgy

-0.62

POSITIVE LOGITS

 protections

0.88

 rights

0.87

 privacy

0.84

 Rights

0.83

 safeguards

0.77

ileged

0.73

 Liberties

0.73

rights

0.73

 Preferences

0.72

eties

0.71

Activations Density 0.021%

privacy-related terms

references to privacy and related policies

No Comments

No Known Activations