INDEX

Explanations

numerical values with a specific format

oai_token-act-pair · gpt-3.5-turbo

numerical values and ratings, often related to measurements, dimensions, or statistical data

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gypt

-0.75

Zen

-0.62

NEO

-0.59

)).

-0.58

idas

-0.58

 Afric

-0.56

lihood

-0.55

 Amen

-0.54

endum

-0.54

zik

-0.53

POSITIVE LOGITS

itol

0.65

umbers

0.56

uple

0.54

ump

0.53

Fas

0.53

oline

0.53

inately

0.52

charging

0.52

 Activate

0.51

ÙĨ

0.51

Activations Density 0.306%

numerical values with a specific format

numerical values and ratings, often related to measurements, dimensions, or statistical data

No Comments

No Known Activations