INDEX

Explanations

mentions of people of color

oai_token-act-pair · gpt-3.5-turbo

references to marginalized groups, specifically people of color

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

Xi

-0.78

 Niet

-0.72

ãĤ´

-0.72

WAR

-0.71

Nex

-0.69

ertodd

-0.68

ERG

-0.68

Sut

-0.68

sg

-0.67

chn

-0.67

POSITIVE LOGITS

blind

0.85

anguage

0.83

 minorities

0.73

 coded

0.69

 backgrounds

0.69

 queer

0.68

 color

0.67

 slurs

0.67

 stripes

0.67

="#

0.65

Activations Density 0.012%

mentions of people of color

references to marginalized groups, specifically people of color

No Comments

No Known Activations