INDEX

Explanations

phrases related to the sensitivity of information

oai_token-act-pair · gpt-3.5-turbo

references to sensitive topics or information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Vale

-0.71

ka

-0.70

leys

-0.69

Fol

-0.68

 Stab

-0.68

 Dale

-0.68

vation

-0.68

 Cities

-0.66

amar

-0.65

 Street

-0.65

POSITIVE LOGITS

 sensitive

3.97

sensitive

2.80

 sensitivity

2.13

 insensitive

2.11

ensitive

2.11

 sensit

1.99

 delicate

1.86

 confidential

1.41

 vulnerable

1.39

 fragile

1.38

Activations Density 0.031%

phrases related to the sensitivity of information

references to sensitive topics or information

No Comments

No Known Activations