INDEX

Explanations

phrases related to errors or issues

oai_token-act-pair · gpt-3.5-turbo

phrases related to user engagement and feedback

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aterasu

-0.67

bane

-0.65

mony

-0.64

forth

-0.64

cknowled

-0.62

warts

-0.62

 foreseeable

-0.60

 Bord

-0.59

birth

-0.57

 sincerely

-0.57

POSITIVE LOGITS

interstitial

0.90

aunder

0.71

currency

0.67

 sear

0.64

%]

0.63

Untitled

0.62

 Search

0.60

sports

0.60

esian

0.59

Brow

0.59

Activations Density 0.086%

phrases related to errors or issues

phrases related to user engagement and feedback

No Comments

No Known Activations