INDEX

Explanations

programming syntax and structure components in code

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

start of user input

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

 Normdatei

-0.51

 TestBed

-0.46

チコミ

-0.45

 BrowserModule

-0.45

fjspx

-0.44

shadowOpacity

-0.43

LayoutPanel

-0.40

gegen

-0.40

Wok

-0.40

 CommonModule

-0.40

POSITIVE LOGITS

Autoritní

0.50

parsedMessage

0.45

Personendaten

0.44

 createState

0.43

WireFormatLite

0.43

Amicalement

0.42

zzleHttp

0.41

 informée

0.41

GOTREF

0.40

 Superan

0.40

Activations Density 0.008%

programming syntax and structure components in code

start of user input

No Comments

No Known Activations