INDEX

Explanations

content recommendations or suggestions

oai_token-act-pair · gpt-3.5-turbo

references to curated selections or recommendations

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

the phrase "Editor's Picks"

oai_token-act-pair · gpt-4-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lich

-0.82

eg

-0.70

sen

-0.68

 Hills

-0.67

lia

-0.66

igi

-0.65

mph

-0.65

ris

-0.65

ev

-0.63

ded

-0.62

POSITIVE LOGITS

 Picks

1.70

ļéĨĴ

1.19

ettings

0.93

ertodd

0.86

CRIP

0.83

ometimes

0.83

pocket

0.83

 Annotations

0.82

uggest

0.81

::::::::

0.77

Activations Density 0.010%

content recommendations or suggestions

references to curated selections or recommendations

the phrase "Editor's Picks"

No Comments

No Known Activations