INDEX

Explanations

words related to challenging or dispelling stereotypes and myths

oai_token-act-pair · gpt-3.5-turbo

references to stereotypes and discussions about their implications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ea

-0.79

 avail

-0.74

 Liquid

-0.71

 liquid

-0.68

ener

-0.66

Aur

-0.65

 pending

-0.65

live

-0.64

pload

-0.64

 authorized

-0.62

POSITIVE LOGITS

 stereotypes

3.57

 stereotype

3.47

 stereotyp

2.67

 stereotypical

2.66

 clich

2.09

 caricature

1.94

 tropes

1.93

 misconceptions

1.78

 prejudices

1.71

 clichÃ©

1.71

Activations Density 0.026%

words related to challenging or dispelling stereotypes and myths

references to stereotypes and discussions about their implications

No Comments

No Known Activations