INDEX

Explanations

numeric values representing statistics or quantities

oai_token-act-pair · gpt-3.5-turbo

the word "just" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 cous

-0.72

 destro

-0.68

undai

-0.67

glomer

-0.66

ĺħ

-0.66

insula

-0.65

 subsequ

-0.65

arella

-0.64

 challeng

-0.63

 PLUS

-0.63

POSITIVE LOGITS

ifiable

1.33

ifications

1.16

IFIED

0.89

IFIC

0.89

ICES

0.89

ices

0.87

ifier

0.80

Enough

0.80

ifiers

0.79

shy

0.78

Activations Density 0.068%

numeric values representing statistics or quantities

the word "just" in various contexts

No Comments

No Known Activations