INDEX

Explanations

references to the media organizations NPR and PBS

oai_token-act-pair · gpt-3.5-turbo

mentions of the NPR (National Public Radio) and PBS (Public Broadcasting Service)

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Spit

-0.68

 Mata

-0.68

wagen

-0.66

gie

-0.65

oldemort

-0.63

 Emirates

-0.63

hire

-0.63

 Haku

-0.62

aneously

-0.62

clave

-0.62

POSITIVE LOGITS

NPR

1.15

NPR

0.99

REC

0.86

 transcripts

0.84

DEF

0.81

OST

0.79

 Transcript

0.77

LV

0.76

 podcasts

0.75

TIT

0.73

Activations Density 0.015%

references to the media organizations NPR and PBS

mentions of the NPR (National Public Radio) and PBS (Public Broadcasting Service)

No Comments

No Known Activations