INDEX

Explanations

law enforcement-related terms and actions

oai_token-act-pair · gpt-3.5-turbo

references to law enforcement personnel

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

FORE

-0.92

ebook

-0.82

nings

-0.78

church

-0.77

rous

-0.76

eful

-0.76

¥µ

-0.74

Virgin

-0.73

hetti

-0.72

lihood

-0.71

POSITIVE LOGITS

 deputies

1.06

 sheriff

1.06

uties

0.88

uty

0.87

 dispatcher

0.87

 confir

0.85

 examiner

0.81

 sher

0.79

 Sheriff

0.79

 colonel

0.78

Activations Density 0.014%

law enforcement-related terms and actions

references to law enforcement personnel

No Comments

No Known Activations