INDEX

Explanations

mentions of toxicity or toxic substances

oai_token-act-pair · gpt-3.5-turbo

references to toxic substances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

hung

-0.79

FORE

-0.78

gain

-0.76

HCR

-0.72

UTERS

-0.71

Bre

-0.71

quart

-0.70

AUT

-0.70

Month

-0.68

ploma

-0.67

POSITIVE LOGITS

 poisoning

1.02

 toxic

0.96

 fumes

0.95

 masculinity

0.94

ologist

0.92

 substances

0.92

oxic

0.90

 waste

0.89

ological

0.89

 substance

0.88

Activations Density 0.013%

mentions of toxicity or toxic substances

references to toxic substances

No Comments

No Known Activations