INDEX

Explanations

phrases related to crime and violence against specific groups or individuals

oai_token-act-pair · gpt-3.5-turbo

references to violence against individuals or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

soDeliveryDate

-0.92

hower

-0.88

gio

-0.77

 helicop

-0.76

BuyableInstoreAndOnline

-0.74

chall

-0.74

chart

-0.73

éĹĺ

-0.72

phrine

-0.70

daq

-0.69

POSITIVE LOGITS

 civilians

1.17

 innocent

1.10

 minorities

0.94

 gays

0.92

 homosexuals

0.91

 humans

0.87

 women

0.86

 foreigners

0.86

 minors

0.83

 humanity

0.83

Activations Density 0.111%

phrases related to crime and violence against specific groups or individuals

references to violence against individuals or groups

No Comments

No Known Activations