INDEX

Explanations

phrases related to inappropriate or offensive content

oai_token-act-pair · gpt-3.5-turbo

language related to offensive or inappropriate content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

UNCH

-0.71

verning

-0.70

OUN

-0.69

sonian

-0.64

ICAL

-0.61

prototype

-0.60

Econom

-0.60

Politics

-0.59

ieth

-0.58

issance

-0.57

POSITIVE LOGITS

 imaginable

0.87

 harming

0.83

 perpetrated

0.81

 arising

0.75

 towards

0.73

 inflicted

0.73

 toward

0.72

 involving

0.71

 affecting

0.71

 emanating

0.71

Activations Density 0.319%

phrases related to inappropriate or offensive content

language related to offensive or inappropriate content

No Comments

No Known Activations