INDEX

Explanations

non-zero values indicating activation signals in technical contexts

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

InitializeComponent()

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

 FetchType

-0.61

enumi

-0.54

tonode

-0.54

 jsPsych

-0.53

 TextInputType

-0.52

delwed

-0.51

MLLoader

-0.49

writeFieldEnd

-0.48

enumii

-0.48

KommentareTeilen

-0.47

POSITIVE LOGITS

contentLoaded

0.56

 CURIAM

0.53

 kasarigan

0.51

UFF

0.47

BibitemOpen

0.47

CodedInputStream

0.43

titious

0.42

зонта

0.41

izados

0.41

 Moment

0.40

Activations Density 0.000%

non-zero values indicating activation signals in technical contexts

InitializeComponent()

No Comments

No Known Activations