INDEX

Explanations

phrases that reference locations or places in time and space

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

the word "in" occurring near certain phrases or contexts

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

eds

-0.65

CVE

-0.64

 chau

-0.63

 inherit

-0.62

alias

-0.61

seless

-0.59

 guiActive

-0.59

maid

-0.57

atel

-0.55

oaded

-0.55

POSITIVE LOGITS

 continuum

0.93

vironment

0.88

 history

0.81

 globe

0.79

between

0.78

animate

0.77

 somew

0.75

 society

0.75

history

0.74

 AMERICA

0.72

Activations Density 0.234%

phrases that reference locations or places in time and space

the word "in" occurring near certain phrases or contexts

No Known Activations