INDEX

Explanations

terms related to different types and contexts of violent behavior, particularly domestic violence

oai_token-act-pair · gpt-3.5-turbo

terms related to domestic violence and abuse

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

*/(

-0.82

Reviewer

-0.74

stanbul

-0.74

aer

-0.73

peror

-0.73

Label

-0.72

travel

-0.71

milo

-0.70

amel

-0.69

source

-0.66

POSITIVE LOGITS

 violence

0.91

 prevention

0.91

 restraining

0.89

 abuse

0.84

 Violence

0.83

 abusive

0.81

 homicides

0.80

 perpetrated

0.80

 suff

0.80

 abusers

0.78

Activations Density 0.038%

terms related to different types and contexts of violent behavior, particularly domestic violence

terms related to domestic violence and abuse

No Comments

No Known Activations