INDEX

Explanations

derogatory language and discriminatory remarks toward individuals or groups

oai_token-act-pair · gpt-3.5-turbo

derogatory language and offensive comments

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

Luck

-0.75

ELY

-0.70

oglu

-0.70

arten

-0.69

 Luck

-0.69

INC

-0.68

ellect

-0.67

ederal

-0.66

 Mechan

-0.65

UNCH

-0.65

POSITIVE LOGITS

 slurs

1.09

 lewd

1.07

 uttered

0.98

 derogatory

0.97

 harassing

0.97

 insulting

0.96

 inappropriately

0.96

 inappropriate

0.94

 indecent

0.93

 homophobic

0.92

Activations Density 0.340%

derogatory language and discriminatory remarks toward individuals or groups

derogatory language and offensive comments

No Comments

No Known Activations