INDEX

Explanations

numerical values and punctuation marks

oai_token-act-pair · gpt-3.5-turbo

representation of code and programming elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 TAMADRA

-0.86

 Union

-0.84

 Laboratory

-0.84

 Smoke

-0.83

 Mobil

-0.82

 Opinion

-0.82

 Marketing

-0.81

 Polk

-0.81

 Investment

-0.80

 Versus

-0.80

POSITIVE LOGITS

cond

1.34

shared

1.30

subject

1.28

apply

1.28

max

1.27

requ

1.27

thanks

1.26

distance

1.26

sorry

1.25

begin

1.25

Activations Density 0.422%

numerical values and punctuation marks

representation of code and programming elements

No Comments

No Known Activations