INDEX

Explanations

code-related elements or structures, particularly focusing on functions and data attributes in programming contexts

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

accessing struct members or array elements

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

 ویکی‌پدی

-0.75

BagLayout

-0.43

 Мексичка

-0.35

 Photocase

-0.34

abstractmethod

-0.34

stateMutability

-0.33

ьаж

-0.31

ImageContext

-0.30

ง

-0.30

UserScript

-0.30

POSITIVE LOGITS

يميديا

0.61

новништво

0.56

exitRule

0.53

DockStyle

0.52

StoreMessageInfo

0.49

 ostavi

0.48

SQLiteDatabase

0.46

PickerController

0.46

HtmlAttribute

0.45

 Administrativna

0.45

Activations Density 0.101%

code-related elements or structures, particularly focusing on functions and data attributes in programming contexts

accessing struct members or array elements

No Comments

No Known Activations