INDEX

Explanations

references to victims in various contexts such as abuse, accidents, or legal cases

oai_token-act-pair · gpt-3.5-turbo

references to victims in various contexts

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

CLASSIFIED

-0.70

OPLE

-0.68

GGGG

-0.66

ove

-0.66

REE

-0.65

 Huck

-0.63

alter

-0.63

leaf

-0.62

yip

-0.62

Style

-0.61

POSITIVE LOGITS

 victims

1.00

 Victims

0.92

vict

0.90

 victim

0.83

izers

0.82

 inflicted

0.80

 Victim

0.80

 Survivors

0.77

hip

0.74

 succumbed

0.74

Activations Density 0.021%

references to victims in various contexts such as abuse, accidents, or legal cases

references to victims in various contexts

No Comments

No Known Activations