INDEX

Explanations

instances where an action should or should not be taken

oai_token-act-pair · gpt-3.5-turbo

phrases indicating obligation or necessity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Wid

-0.73

 Cros

-0.69

 Wick

-0.68

 Vaughan

-0.65

 Suzuki

-0.63

 theorem

-0.61

Fra

-0.61

 Hanson

-0.60

 Dahl

-0.60

 Weeks

-0.60

POSITIVE LOGITS

fitting

1.10

 able

1.05

 ashamed

1.03

 judged

1.01

 treated

0.99

 regarded

0.98

 viewed

0.97

 avoided

0.97

 applauded

0.97

 considered

0.94

Activations Density 0.114%

instances where an action should or should not be taken

phrases indicating obligation or necessity

No Comments

No Known Activations