INDEX

Explanations

instances where information is revealed or made known to others

oai_token-act-pair · gpt-3.5-turbo

instances of the word "disclose" and its variations

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

âĿ

-0.72

 Adolf

-0.69

 Statue

-0.66

 Anim

-0.65

 memes

-0.64

jo

-0.64

 Animated

-0.64

 zombie

-0.64

Pol

-0.63

 Luck

-0.62

POSITIVE LOGITS

 disclosed

3.16

 disclose

3.10

 disclosing

2.67

 disclosure

2.49

 discl

2.46

 disclosures

2.38

 Disclosure

1.87

 divul

1.86

 undisclosed

1.21

closure

1.20

Activations Density 0.016%

instances where information is revealed or made known to others

instances of the word "disclose" and its variations

No Comments

No Known Activations