INDEX

Explanations

religious references or contexts in which authority figures are speaking

oai_token-act-pair · gpt-3.5-turbo

phrases related to violent or aggressive actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 homebrew

-0.87

 Wilmington

-0.82

 Downs

-0.79

 chirop

-0.79

 Staten

-0.78

 Craigslist

-0.77

 downs

-0.77

 brunch

-0.76

 Franklin

-0.75

 Asheville

-0.75

POSITIVE LOGITS

Ø§ÙĦ

1.47

ÙĪ

1.43

Ø

1.37

Ø§Ø

1.36

Ù

1.36

Ø§

1.34

Islamic

1.34

Øª

1.34

Pakistan

1.33

Ùħ

1.32

Activations Density 0.360%

religious references or contexts in which authority figures are speaking

phrases related to violent or aggressive actions

No Comments

No Known Activations