INDEX

Explanations

phrases related to things being well-defined or well-described

oai_token-act-pair · gpt-3.5-turbo

phrases indicating something that is well-defined or well-regarded

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

 Midnight

-0.71

 Cutter

-0.70

Phi

-0.69

 Bravo

-0.69

 Crisis

-0.69

 Compass

-0.67

 Theft

-0.67

 Indigo

-0.66

 Tags

-0.64

 Hancock

-0.64

POSITIVE LOGITS

known

1.37

established

1.33

defined

1.30

enough

1.28

trained

1.27

earned

1.26

documented

1.25

connected

1.24

respected

1.23

intention

1.23

Activations Density 0.032%

phrases related to things being well-defined or well-described

phrases indicating something that is well-defined or well-regarded

No Comments

No Known Activations