INDEX

Explanations

class and method definitions in programming code

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

App main program entry point

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

ftagPool

-0.98

 Chwiliwch

-0.93

 otomatig

-0.89

AndEndTag

-0.83

IsMutable

-0.71

፩

-0.71

 propOrder

-0.70

 يتيمه

-0.68

 oprot

-0.68

oa̍t

-0.67

POSITIVE LOGITS

 Main

0.74

Main

0.74

MAIN

0.50

 AppComponent

0.48

main

0.47

 MAIN

0.43

getMain

0.39

UnitTesting

0.35

App

0.33

 main

0.31

Activations Density 0.001%

class and method definitions in programming code

App main program entry point

No Comments

No Known Activations