INDEX

Explanations

references to a popular culture phenomenon called "Pop" with varying levels of relatedness

oai_token-act-pair · gpt-3.5-turbo

references to pop culture, especially related to "Pop" or "Pop-up" events and phenomena

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

OME

-0.69

beit

-0.63

IVES

-0.62

Ket

-0.59

RAW

-0.58

ILD

-0.57

 archived

-0.57

 Parameters

-0.53

 Leth

-0.53

GW

-0.53

POSITIVE LOGITS

corn

1.37

ulates

1.37

ulating

1.36

ulations

1.34

ularity

1.31

ular

1.29

ulated

1.29

ULAR

1.13

ulous

1.13

ulate

1.13

Activations Density 0.027%

references to a popular culture phenomenon called "Pop" with varying levels of relatedness

references to pop culture, especially related to "Pop" or "Pop-up" events and phenomena

No Comments

No Known Activations