INDEX

Explanations

elements related to user interface text in programming code

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

labels followed by data

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

 betweenstory

-0.57

 intStringLen

-0.54

TokenNameDOT

-0.49

الإنجليزية

-0.49

فایل‌لار

-0.49

الحياه

-0.49

vège

-0.47

IonicModule

-0.46

 OMITBAD

-0.46

 condi

-0.45

POSITIVE LOGITS

 nahilalakip

0.61

TabIndex

0.54

UseVisualStyle

0.46

 مشين

0.46

setVerticalGroup

0.46

VIAF

0.43

StoryboardSegue

0.42

fjspx

0.41

GOTREF

0.40

AccessorTable

0.37

Activations Density 0.000%

elements related to user interface text in programming code

labels followed by data

No Comments

No Known Activations