INDEX

Explanations

privacy-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to privacy and security

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

terday

-0.86

kell

-0.82

leys

-0.71

uana

-0.70

gling

-0.67

 McDonnell

-0.66

steen

-0.65

conom

-0.64

 Toad

-0.64

manship

-0.63

POSITIVE LOGITS

 preferences

0.79

Privacy

0.73

 preference

0.73

 actionGroup

0.69

iframe

0.67

policy

0.66

 privacy

0.66

protected

0.65

rights

0.64

 intrusive

0.64

Activations Density 0.020%

privacy-related terms

terms related to privacy and security

No Comments

No Known Activations