INDEX

Explanations

mentions of crimes like murder

oai_token-act-pair · gpt-3.5-turbo

instances of the word "murder."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Cola

-0.80

wcsstore

-0.76

ais

-0.73

UTC

-0.73

BuyableInstoreAndOnline

-0.70

Dub

-0.68

imus

-0.67

uve

-0.67

ube

-0.66

arity

-0.65

POSITIVE LOGITS

 spree

1.15

 murder

1.03

 murders

0.93

ously

0.90

 homicide

0.89

hyde

0.87

 rampage

0.86

 murderer

0.86

 murdering

0.85

 Murder

0.84

Activations Density 0.031%

mentions of crimes like murder

instances of the word "murder."

No Comments

No Known Activations