INDEX

Explanations

references to various acronyms or abbreviations that include the letters "MC."

oai_token-act-pair · gpt-3.5-turbo

mentions of "MC" or related acronyms and terms associated with it

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Aval

-0.69

ãĤ¯

-0.67

Ô

-0.67

 Alaska

-0.66

*/(

-0.65

ãĥİ

-0.64

 sane

-0.63

 Gingrich

-0.63

éĹĺ

-0.62

ãĥ³ãĤ¸

-0.62

POSITIVE LOGITS

MC

1.60

Fs

0.95

ascade

0.85

abre

0.84

illion

0.82

SA

0.81

CC

0.80

MP

0.80

SC

0.79

ADA

0.78

Activations Density 0.007%

references to various acronyms or abbreviations that include the letters "MC."

mentions of "MC" or related acronyms and terms associated with it

No Comments

No Known Activations