INDEX

Explanations

text related to instructions or information

oai_token-act-pair · gpt-3.5-turbo

terms related to surveys and questionnaires

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

forth

-0.69

 beforehand

-0.54

eering

-0.54

john

-0.52

ably

-0.51

lihood

-0.49

 undet

-0.49

 writ

-0.49

lessness

-0.48

resp

-0.47

POSITIVE LOGITS

 ILCS

0.78

quickShipAvailable

0.60

BuyableInstoreAndOnline

0.60

apter

0.57

earances

0.57

ilogy

0.57

istar

0.56

umbn

0.56

\<

0.56

chemist

0.55

Activations Density 0.172%

text related to instructions or information

terms related to surveys and questionnaires

No Comments

No Known Activations