INDEX

Explanations

the word "demon" or variations of it

oai_token-act-pair · gpt-3.5-turbo

terms related to demonization and stigmatization of groups or individuals

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

etry

-0.93

 Newport

-0.88

 Howe

-0.75

 Pearce

-0.75

borough

-0.73

 Howell

-0.71

 Wiggins

-0.71

 Atkinson

-0.70

 Cargo

-0.69

 Webb

-0.68

POSITIVE LOGITS

 demon

3.76

 demons

2.60

demon

2.38

 demonic

2.05

Demon

1.97

 Demon

1.90

 Demons

1.67

 exorc

1.62

 Demonic

1.62

 devil

1.35

Activations Density 0.021%

the word "demon" or variations of it

terms related to demonization and stigmatization of groups or individuals

No Comments

No Known Activations