INDEX

Explanations

the word "are" followed by a word

oai_token-act-pair · gpt-3.5-turbo

the phrase "you are" 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

osate

-0.82

ESE

-0.67

uish

-0.64

ð

-0.64

FY

-0.63

ionics

-0.62

udeau

-0.61

 Restore

-0.60

 Rove

-0.60

 suffice

-0.60

POSITIVE LOGITS

 gonna

0.93

nt

0.93

 able

0.89

 yourself

0.85

 lucky

0.83

 choosing

0.80

 intimately

0.79

 willing

0.79

 supposed

0.78

 going

0.78

Activations Density 0.139%

the word "are" followed by a word

the phrase "you are" in various contexts

No Comments

No Known Activations