INDEX

Explanations

comparisons or numerical figures presenting quantities being greater than or less than a reference value

oai_token-act-pair · gpt-3.5-turbo

comparative phrases indicating quantities or measurements

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

ard

-0.68

 seriousness

-0.66

 authenticity

-0.65

 direction

-0.65

culosis

-0.64

 refrain

-0.64

RAW

-0.63

 restriction

-0.63

OTA

-0.62

MT

-0.62

POSITIVE LOGITS

average

0.80

 average

0.78

whelming

0.73

 Â±

0.70

mean

0.68

whel

0.67

=>

0.67

 expected

0.66

 Average

0.66

agles

0.64

Activations Density 0.105%

comparisons or numerical figures presenting quantities being greater than or less than a reference value

comparative phrases indicating quantities or measurements

No Comments

No Known Activations