INDEX

Explanations

numbers related to quantities or statistics

oai_token-act-pair · gpt-3.5-turbo

numerical figures and statistics

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

ĸļ

-0.69

 Gemini

-0.69

 rematch

-0.66

ModLoader

-0.65

 Mystic

-0.64

 symmetry

-0.64

 Accountability

-0.64

 Bomber

-0.63

 longevity

-0.63

 Shal

-0.63

POSITIVE LOGITS

 dozen

0.91

 thousand

0.90

 hundred

0.89

million

0.85

uously

0.83

illion

0.82

venth

0.82

eteenth

0.81

dozen

0.78

 million

0.75

Activations Density 0.069%

numbers related to quantities or statistics

numerical figures and statistics

No Comments

No Known Activations