INDEX

Explanations

mentions of names or possibly entities in a given context

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and their connections or relationships

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

afety

-0.63

pmwiki

-0.60

iquette

-0.56

 FOIA

-0.56

DragonMagazine

-0.52

 shady

-0.51

 trustworthy

-0.51

 Breach

-0.51

 investig

-0.50

 confidentiality

-0.50

POSITIVE LOGITS

Jr

0.66

lda

0.64

chev

0.63

inki

0.61

ovych

0.60

etter

0.57

ophers

0.53

enei

0.52

bye

0.52

venge

0.52

Activations Density 1.437%

mentions of names or possibly entities in a given context

references to specific individuals and their connections or relationships

No Comments

No Known Activations