INDEX

Explanations

phrases related to reading continuation prompts in articles

oai_token-act-pair · gpt-3.5-turbo

references to the act of reading

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

»Ĵ

-0.77

ignt

-0.68

 Gleaming

-0.67

 Compan

-0.64

ĪĴ

-0.63

 Profession

-0.62

 Gathering

-0.61

Ã±o

-0.58

ocally

-0.57

esville

-0.57

POSITIVE LOGITS

 Below

0.66

...]

0.65

 BELOW

0.64

 toggle

0.62

 aloud

0.60

 âĨĴ

0.59

ILY

0.59

wheel

0.57

 ARTICLE

0.57

theless

0.57

Activations Density 0.013%

phrases related to reading continuation prompts in articles

references to the act of reading

No Comments

No Known Activations