INDEX

Explanations

references to programming concepts and components related to data processing

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

code, names, and variables

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Italijanski

-0.64

AsUp

-0.58

 propOrder

-0.57

InjectAttribute

-0.56

 CanadaChoose

-0.53

 ſind

-0.53

setVerticalGroup

-0.53

 &___

-0.52

EndGlobalSection

-0.51

parsedMessage

-0.51

POSITIVE LOGITS

WriteBarrier

0.40

 mergeFrom

0.33

///</

0.33

Géographie

0.32

 Genuss

0.32

 intptr

0.31

NullException

0.31

fvar

0.31

 wide

0.30

ungszeit

0.30

Activations Density 0.064%

references to programming concepts and components related to data processing

code, names, and variables

No Comments

No Known Activations