INDEX

Explanations

terms and concepts related to the training process of artificial neural networks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

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

jā

-0.39

 hetzelfde

-0.38

 Schmitz

-0.35

 afges

-0.34

 łat

-0.34

 fonti

-0.33

 jabatan

-0.33

缺少

-0.33

Cyfeiriadau

-0.32

应

-0.32

POSITIVE LOGITS

'{@

0.53

 <=",

0.53

$_['

0.50

➯

0.50

postsleuth

0.49

endphp

0.48

oredCriteria

0.47

 Learning

0.47

 Epoch

0.45

 ―――――

0.45

Activations Density 0.014%

terms and concepts related to the training process of artificial neural networks

No Comments

No Known Activations