INDEX

Explanations

verbs related to performance or entertainment industries, especially acting

oai_token-act-pair · gpt-3.5-turbo

references to acting and performances

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

-0.78

avez

-0.76

imus

-0.73

etsk

-0.72

 opio

-0.71

ascus

-0.68

£ı

-0.67

unin

-0.66

ebin

-0.65

ptoms

-0.65

POSITIVE LOGITS

 actress

0.91

 studio

0.90

 actresses

0.88

 studios

0.84

 arts

0.84

 gigs

0.84

writing

0.78

 audition

0.77

anship

0.75

Actor

0.75

Activations Density 0.368%

verbs related to performance or entertainment industries, especially acting

references to acting and performances

No Comments

No Known Activations