INDEX

Explanations

action words related to creating or making one's own something

oai_token-act-pair · gpt-3.5-turbo

references to creating or making something personal or customized

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

anwhile

-0.87

Mub

-0.80

hovah

-0.72

EGIN

-0.69

ynchronous

-0.64

trump

-0.64

ortunately

-0.64

isSpecialOrderable

-0.64

hesda

-0.63

lehem

-0.63

POSITIVE LOGITS

 custom

0.87

 version

0.85

 customized

0.79

 personal

0.77

 homegrown

0.74

 creations

0.74

 unique

0.73

 selves

0.72

 personalized

0.71

 copy

0.71

Activations Density 0.035%

action words related to creating or making one's own something

references to creating or making something personal or customized

No Comments

No Known Activations