INDEX

Explanations

names of hotels or places with the word "vivort" or variations of it

oai_token-act-pair · gpt-3.5-turbo

frequency of specific syllabic patterns and letter combinations

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

inki

-0.66

 lest

-0.63

SpaceEngineers

-0.63

 shove

-0.60

 withd

-0.60

wegian

-0.59

bender

-0.59

 bible

-0.56

ghan

-0.56

 wound

-0.56

POSITIVE LOGITS

bilt

0.96

export

0.80

ovsky

0.71

achev

0.68

tsky

0.68

opa

0.67

wagen

0.66

arde

0.65

idis

0.65

vich

0.63

Activations Density 0.306%

names of hotels or places with the word "vivort" or variations of it

frequency of specific syllabic patterns and letter combinations

No Comments

No Known Activations