INDEX

Explanations

words related to negative behaviors or consequences such as hatred, prejudice, harassment, violence, and rejection

oai_token-act-pair · gpt-3.5-turbo

references to systemic discrimination and social injustices

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

ufact

-0.72

çīĪ

-0.66

eus

-0.66

arten

-0.63

cko

-0.62

eport

-0.61

ppa

-0.60

utra

-0.58

arers

-0.58

illac

-0.57

POSITIVE LOGITS

 imprisonment

0.70

 unwanted

0.70

 harassment

0.66

 boredom

0.65

 threats

0.65

vous

0.65

 loneliness

0.63

 deaths

0.63

 bullying

0.62

 misinformation

0.62

Activations Density 0.474%

words related to negative behaviors or consequences such as hatred, prejudice, harassment, violence, and rejection

references to systemic discrimination and social injustices

No Comments

No Known Activations