INDEX

Explanations

proper nouns related to individuals named "Kat" or variations of the name

oai_token-act-pair · gpt-3.5-turbo

references to individuals with the name "Kat" or variations of it

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

stood

-0.70

nesota

-0.65

purpose

-0.64

dayName

-0.63

cloth

-0.63

ences

-0.61

character

-0.60

uploads

-0.59

 mistaken

-0.58

ministic

-0.57

POSITIVE LOGITS

rina

1.18

apult

1.02

arine

0.96

rine

0.94

Von

0.94

apesh

0.93

zman

0.93

zen

0.90

mand

0.89

usha

0.87

Activations Density 0.038%

proper nouns related to individuals named "Kat" or variations of the name

references to individuals with the name "Kat" or variations of it

No Comments

No Known Activations