INDEX

Explanations

mentions of objects or entities that are described as being of a certain size

oai_token-act-pair · gpt-3.5-turbo

references to different size categories

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

 Dial

-0.65

 Anarchy

-0.61

 confer

-0.61

 revel

-0.61

 general

-0.60

 Vine

-0.59

Man

-0.59

 Stead

-0.59

 Church

-0.58

 reve

-0.58

POSITIVE LOGITS

sized

3.85

size

2.84

 sized

2.40

shaped

1.59

Size

1.52

SIZE

1.51

length

1.49

scale

1.44

 size

1.39

 sizes

1.28

Activations Density 0.017%

mentions of objects or entities that are described as being of a certain size

references to different size categories

No Comments

No Known Activations