INDEX

Explanations

programming-related terms and structures, particularly those associated with class and method definitions

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

programming extension methods

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

ListItemIcon

-0.34

 vaih

-0.34

="@+

-0.33

TCHAR

-0.32

WebpackPlugin

-0.31

 beginnetje

-0.31

awtextra

-0.31

 насељу

-0.31

 Irán

-0.31

 liitty

-0.30

POSITIVE LOGITS

Extensions

0.63

 functools

0.59

 Extensions

0.59

jspx

0.58

MigrationBuilder

0.57

Хьажоргаш

0.56

 extensions

0.55

 frumos

0.53

ScopeManager

0.53

IndentedString

0.53

Activations Density 0.007%

programming-related terms and structures, particularly those associated with class and method definitions

programming extension methods

No Comments

No Known Activations