INDEX

Explanations

phrases related to social groups or societies

oai_token-act-pair · gpt-3.5-turbo

mentions of societal themes or concepts related to society

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Args

-0.86

 Miranda

-0.74

uggage

-0.73

ainer

-0.73

 Forrest

-0.71

uru

-0.70

ofer

-0.67

src

-0.65

annon

-0.64

brid

-0.63

POSITIVE LOGITS

Soc

3.26

 societies

2.85

Soc

2.40

soc

2.13

 Society

1.90

SOC

1.67

soc

1.65

 Soci

1.56

 society

1.46

ocial

1.34

Activations Density 0.024%

phrases related to social groups or societies

mentions of societal themes or concepts related to society

No Comments

No Known Activations