INDEX

Explanations

numbers mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "number" followed by numerical values

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

SPONSORED

-0.86

kes

-0.73

itars

-0.69

ende

-0.67

TOR

-0.66

 VIDEOS

-0.63

rained

-0.61

endez

-0.61

pret

-0.60

Äĩ

-0.60

POSITIVE LOGITS

 number

3.53

number

2.63

 Number

2.35

Number

2.30

 numbers

2.26

 Numbers

1.92

 amount

1.80

NUM

1.76

 percentage

1.54

 numbered

1.54

Activations Density 0.039%

numbers mentioned in the text

occurrences of the word "number" followed by numerical values

No Comments

No Known Activations