INDEX

Explanations

mentions of discrimination based on various factors, such as sexual orientation, race, and disability

oai_token-act-pair · gpt-3.5-turbo

references to discrimination, particularly in legal or social contexts

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

ski

-0.72

Adds

-0.71

cycle

-0.71

tom

-0.70

TOR

-0.68

DCS

-0.68

bold

-0.67

hran

-0.66

sis

-0.66

links

-0.65

POSITIVE LOGITS

 discrimination

0.97

rimination

0.92

 prejudice

0.85

 discriminated

0.84

 Discrimination

0.82

 protections

0.79

 retaliation

0.79

 slurs

0.78

 prejud

0.78

 discriminating

0.77

Activations Density 0.034%

mentions of discrimination based on various factors, such as sexual orientation, race, and disability

references to discrimination, particularly in legal or social contexts

No Comments

No Known Activations