INDEX

Explanations

phrases related to violent incidents or criminal activities involving death

oai_token-act-pair · gpt-3.5-turbo

phrases related to deaths in various contexts and their implications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

eele

-0.83

wcsstore

-0.79

Developer

-0.78

efully

-0.76

 explan

-0.73

URI

-0.72

chart

-0.72

hern

-0.71

 CLSID

-0.71

arak

-0.69

POSITIVE LOGITS

 innocent

0.78

thood

0.74

 unborn

0.74

 whistle

0.71

 innocence

0.69

 Debbie

0.68

 deceased

0.68

Mum

0.66

 Jill

0.64

 spree

0.63

Activations Density 0.095%

phrases related to violent incidents or criminal activities involving death

phrases related to deaths in various contexts and their implications

No Comments

No Known Activations