INDEX

Explanations

references to historical artifacts or concepts, specifically related to the term "Ark."

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Ark."

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

ãģ¦

-0.91

phrine

-0.73

BOOK

-0.71

cussion

-0.69

ãģŁ

-0.68

bered

-0.68

payer

-0.66

ãĥ¼ãĥ³

-0.66

 Gadget

-0.62

GROUP

-0.62

POSITIVE LOGITS

ansas

1.49

adia

1.01

angel

1.00

 Encounter

0.99

wright

0.98

adelphia

0.96

arium

0.95

esian

0.94

alis

0.92

reath

0.89

Activations Density 0.012%

references to historical artifacts or concepts, specifically related to the term "Ark."

mentions of the term "Ark."

No Comments

No Known Activations