INDEX

Explanations

sequences of technical terms and parameters related to data processing and APIs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ROUTINE

-0.54

 végétale

-0.53

encodeWith

-0.50

 cerevisiae

-0.50

 Darlington

-0.50

ientôt

-0.49

enumi

-0.48

 conséquence

-0.47

λίου

-0.47

 slightest

-0.47

POSITIVE LOGITS

 mergeFrom

0.60

λογία

0.58

ագրություններ

0.58

aarrggbb

0.56

例句

0.55

mobileqq

0.52

ledem

0.51

 opérateur

0.50

SharedCtor

0.50

Begriffsklä

0.48

Activations Density 0.022%

sequences of technical terms and parameters related to data processing and APIs

No Comments

No Known Activations