INDEX

Explanations

symbols and numbers in a specific format

oai_token-act-pair · gpt-3.5-turbo

instances of numeric values or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Mub

-0.81

 crocod

-0.81

 Robot

-0.79

 Haku

-0.77

 Tier

-0.77

 Shia

-0.76

Ide

-0.75

Sob

-0.74

 Solomon

-0.73

Syd

-0.73

POSITIVE LOGITS

matter

1.34

together

1.33

enough

1.32

Page

1.32

Reviewer

1.31

above

1.27

their

1.26

appropriate

1.25

older

1.25

that

1.25

Activations Density 0.119%

symbols and numbers in a specific format

instances of numeric values or measurements

No Comments

No Known Activations