INDEX

Explanations

references to readers or audience in written content

oai_token-act-pair · gpt-3.5-turbo

references to readers and their involvement or opinions

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

 Guan

-0.64

 Romo

-0.63

 Pharma

-0.62

ering

-0.61

 Bain

-0.60

 Skull

-0.60

 Sabb

-0.59

 Gutierrez

-0.59

 Vulcan

-0.58

 Mandal

-0.58

POSITIVE LOGITS

hip

1.40

 reader

1.00

 readers

0.96

uggest

0.89

 Readers

0.89

DragonMagazine

0.88

pace

0.84

0.82

reader

0.82

 Reader

0.82

Activations Density 0.016%

references to readers or audience in written content

references to readers and their involvement or opinions

No Comments

No Known Activations