INDEX

Explanations

mentions of processes and investigations in technical reports or academic discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

舺

-0.47

TestingModule

-0.46

++]

-0.45

oa̍t

-0.45

 getItemCount

-0.45

verwijspagina

-0.43

tangentMode

-0.42

 вікісторінку

-0.41

ResumeLayout

-0.40

 bezeichneter

-0.40

POSITIVE LOGITS

addContainerGap

0.47

RefNanny

0.46

tagHelperRunner

0.45

 समीक्षक

0.41

<?

0.40

 Stampa

0.40

Pork

0.39

RTLR

0.39

 starting

0.38

AlterField

0.38

Activations Density 0.771%

mentions of processes and investigations in technical reports or academic discussions

No Comments

No Known Activations