INDEX

Explanations

references to acts of killing or violence

oai_token-act-pair · gpt-3.5-turbo

the word "Kill" in various contexts

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

tem

-0.79

 contracting

-0.72

 provisional

-0.72

 prism

-0.71

rup

-0.71

 plane

-0.69

abb

-0.68

 cred

-0.67

notation

-0.67

 notation

-0.66

POSITIVE LOGITS

 Kill

3.82

Kill

2.78

 Killing

1.98

 KILL

1.90

 Kills

1.83

kill

1.83

 Killer

1.62

 Slaughter

1.53

 Murd

1.52

 kill

1.52

Activations Density 0.021%

references to acts of killing or violence

the word "Kill" in various contexts

No Comments

No Known Activations