INDEX

Explanations

mentions of discriminatory acts based on various factors like age, race, gender, and sexual orientation

oai_token-act-pair · gpt-3.5-turbo

references to discrimination across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

annis

-0.77

Adds

-0.74

orah

-0.71

bold

-0.69

Pod

-0.67

tom

-0.66

kos

-0.66

ski

-0.66

Interstitial

-0.65

sis

-0.65

POSITIVE LOGITS

 discrimination

1.16

 Discrimination

0.98

 prejudice

0.94

 discriminated

0.93

rimination

0.90

yip

0.87

discrimination

0.87

 protections

0.85

 slurs

0.85

 discriminate

0.80

Activations Density 0.016%

mentions of discriminatory acts based on various factors like age, race, gender, and sexual orientation

references to discrimination across various contexts

No Comments

No Known Activations