INDEX

Explanations

instances where the word "more" is followed by a number

oai_token-act-pair · gpt-3.5-turbo

references to increasing quantities or numbers of people or items

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

xtap

-0.86

imaru

-0.83

ĺħ

-0.82

ishops

-0.70

DragonMagazine

-0.69

 Conclusion

-0.67

translation

-0.67

Fram

-0.67

Gra

-0.66

Guard

-0.65

POSITIVE LOGITS

 than

1.29

 stringent

1.02

 sophisticated

0.89

ado

0.88

 elaborate

0.84

 importantly

0.84

than

0.83

 frequent

0.82

 Than

0.81

 extensive

0.79

Activations Density 0.153%

instances where the word "more" is followed by a number

references to increasing quantities or numbers of people or items

No Comments

No Known Activations