INDEX

Explanations

numeric ranges presented in a specific format

oai_token-act-pair · gpt-3.5-turbo

negative or critical language, particularly concerning statistics and assessments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 turbulence

-0.71

 Cassidy

-0.70

Cav

-0.68

 Fiesta

-0.68

AVG

-0.65

 Brav

-0.64

 Aether

-0.64

 Collider

-0.63

NP

-0.62

NV

-0.61

POSITIVE LOGITS

five

1.52

six

1.50

eight

1.49

seven

1.48

nine

1.46

four

1.38

three

1.30

two

1.28

fifth

1.23

undreds

1.16

Activations Density 0.033%

numeric ranges presented in a specific format

negative or critical language, particularly concerning statistics and assessments

No Comments

No Known Activations