INDEX

Explanations

phrases related to causing harm or violence

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "take out"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

gnu

-0.72

ould

-0.70

ogue

-0.69

aug

-0.68

ndra

-0.68

isma

-0.67

Ã¥

-0.65

Values

-0.64

ept

-0.64

 Mysteries

-0.63

POSITIVE LOGITS

stretched

0.79

 swat

0.68

 trash

0.66

 weeds

0.66

ãĥīãĥ©

0.64

 loans

0.64

 havoc

0.64

 toll

0.62

otom

0.62

shot

0.61

Activations Density 0.042%

phrases related to causing harm or violence

instances of the phrase "take out"

No Comments

No Known Activations