INDEX

Explanations

comparative words and phrases, specifically focusing on "more" in this case

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize additional information or emphasis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

xtap

-0.79

atan

-0.67

uca

-0.66

oresc

-0.65

owe

-0.64

 Classification

-0.63

OCK

-0.62

-0.61

DragonMagazine

-0.61

 Ward

-0.61

POSITIVE LOGITS

 importantly

1.32

 than

1.31

ado

0.87

 stringent

0.87

 Than

0.85

 expensive

0.83

ened

0.83

 elaborate

0.81

 likely

0.80

 efficient

0.80

Activations Density 0.146%

comparative words and phrases, specifically focusing on "more" in this case

phrases that emphasize additional information or emphasis

No Comments

No Known Activations