INDEX

Explanations

technical terms related to data handling and system architecture

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

ftagPool

-0.44

"?>

-0.40

LookAnd

-0.39

 preg

-0.38

/**

-0.36

GetBytes

-0.34

UnitTesting

-0.34

//

-0.33

הערות

-0.33

POSITIVE LOGITS

 fieltro

0.64

adaptiveStyles

0.63

 artificiales

0.63

 noDo

0.59

exitRule

0.53

corazón

0.52

 meditación

0.52

 húmedo

0.51

cartão

0.51

 feroit

0.51

Activations Density 0.012%

technical terms related to data handling and system architecture

No Comments

No Known Activations