INDEX

Explanations

phrases related to domestic violence and work-related health issues

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

Data Type

torch.float32

Hook Point

blocks.8.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

ĸļ

-0.92

ometry

-0.82

itone

-0.79

Tokens

-0.78

iture

-0.77

ulously

-0.76

utical

-0.76

ometer

-0.76

ruary

-0.76

 largeDownload

-0.74

POSITIVE LOGITS

 disease

1.46

 violence

1.37

 corruption

1.34

 terrorism

1.34

 diseases

1.31

 infection

1.31

 crime

1.30

 piracy

1.29

 vandalism

1.29

 theft

1.28

Activations Density 4.125%

phrases related to domestic violence and work-related health issues

No Comments

No Known Activations