INDEX

Explanations

information about specific topics or statistical data

oai_token-act-pair · gpt-3.5-turbo

repeated phrases or patterns in statements

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

cru

-0.70

 wound

-0.63

eries

-0.63

 rush

-0.62

 engagement

-0.61

 stand

-0.60

nor

-0.59

ery

-0.59

 grand

-0.57

endor

-0.56

POSITIVE LOGITS

 Some

2.82

Some

2.18

 Many

1.84

 SOME

1.80

 Others

1.79

 Sometimes

1.77

 Several

1.73

some

1.68

 Certain

1.60

 Occasionally

1.57

Activations Density 0.022%

information about specific topics or statistical data

repeated phrases or patterns in statements

No Comments

No Known Activations