INDEX

Explanations

statements where someone is deemed to be correct or accurate

oai_token-act-pair · gpt-3.5-turbo

phrases asserting correctness or validity

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

 Flavoring

-0.83

 Gong

-0.76

ĸļ

-0.72

 scrim

-0.69

cano

-0.69

WAYS

-0.68

andom

-0.67

ility

-0.66

 Pastebin

-0.64

 Remastered

-0.62

POSITIVE LOGITS

eous

1.00

fully

0.84

headed

0.80

 smack

0.79

sighted

0.77

footed

0.75

Osw

0.74

wing

0.73

 fielder

0.72

eyed

0.71

Activations Density 0.035%

statements where someone is deemed to be correct or accurate

phrases asserting correctness or validity

No Comments

No Known Activations