INDEX

Explanations

words related to bias, discrimination, and negative preconceived notions about certain groups of people

oai_token-act-pair · gpt-3.5-turbo

terms related to prejudice and discrimination

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Interstitial

-0.85

VID

-0.81

sis

-0.76

adra

-0.75

avez

-0.75

ascus

-0.74

incinn

-0.73

ODE

-0.72

irgin

-0.70

ramid

-0.70

POSITIVE LOGITS

 prejudice

1.28

 prejud

1.20

 prejudices

0.95

eering

0.83

icial

0.78

 hatred

0.78

 intolerance

0.74

ophobic

0.74

wart

0.73

 towards

0.70

Activations Density 0.011%

words related to bias, discrimination, and negative preconceived notions about certain groups of people

terms related to prejudice and discrimination

No Comments

No Known Activations