INDEX

Explanations

references to toxic substances or materials

oai_token-act-pair · gpt-3.5-turbo

references to toxic substances and their harmful effects

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

Finish

-0.78

iture

-0.76

Enjoy

-0.76

Joined

-0.71

ï¿½ï¿½

-0.70

 Arri

-0.69

Gim

-0.69

 Bauer

-0.69

LU

-0.68

 Annotations

-0.67

POSITIVE LOGITS

 toxic

3.57

 Toxic

2.37

 poisonous

2.33

 toxicity

2.29

oxic

2.03

 toxins

1.93

 toxin

1.79

 corros

1.73

 poison

1.73

 neurot

1.66

Activations Density 0.023%

references to toxic substances or materials

references to toxic substances and their harmful effects

No Comments

No Known Activations