INDEX

Explanations

instances of phrases related to a wide range of options, characteristics, or categories

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a variety of options or tasks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

iquette

-0.68

boxing

-0.62

 Statue

-0.61

 nearest

-0.61

 Burke

-0.59

 Lumpur

-0.59

irs

-0.58

riers

-0.58

prints

-0.57

 Quote

-0.57

POSITIVE LOGITS

 sorts

0.89

course

0.77

 course

0.76

 adject

0.66

vari

0.66

 topics

0.65

 ways

0.64

luster

0.62

 disparate

0.61

 kindred

0.61

Activations Density 0.219%

instances of phrases related to a wide range of options, characteristics, or categories

phrases indicating a variety of options or tasks

No Comments

No Known Activations