INDEX

Explanations

numerical quantities such as 'hundreds' or 'thousands'

oai_token-act-pair · gpt-3.5-turbo

quantifiers that denote large quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

inventoryQuantity

-0.57

agate

-0.56

 aggression

-0.55

 omission

-0.54

iban

-0.54

estine

-0.53

 isolation

-0.53

 goodness

-0.53

ANI

-0.52

iator

-0.52

POSITIVE LOGITS

ourcing

1.12

of

1.06

ourced

1.01

paces

0.96

 upon

0.92

 thereof

0.89

ynthesis

0.79

Of

0.78

 thousand

0.78

chool

0.76

Activations Density 0.045%

numerical quantities such as 'hundreds' or 'thousands'

quantifiers that denote large quantities

No Comments

No Known Activations