INDEX

Explanations

terms related to discrimination and its various forms

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

phrases related to discrimination based on various characteristics

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ECD

-0.76

Interstitial

-0.76

sis

-0.72

bold

-0.70

ski

-0.68

Pod

-0.67

Sed

-0.67

efer

-0.67

-0.66

Phones

-0.66

POSITIVE LOGITS

 discrimination

1.36

 Discrimination

1.18

 discriminated

1.09

 prejudice

1.03

 discriminate

0.99

 discriminating

0.95

 discrim

0.94

 discriminatory

0.91

 prejudices

0.91

 bigotry

0.90

Activations Density 0.018%

terms related to discrimination and its various forms

phrases related to discrimination based on various characteristics

No Comments

No Known Activations