INDEX

Explanations

programming-related keywords and code structures

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

code-related identifiers

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

ViewFeatures

-0.39

cin

-0.38

CopyWith

-0.33

">//

-0.33

ment

-0.32

oredCriteria

-0.32

$=

-0.32

sm

-0.31

setupUi

-0.30

 znajdu

-0.30

POSITIVE LOGITS

లాలు

0.55

WriteTagHelper

0.54

tagHelperRunner

0.54

 OkHttpClient

0.53

 للاسماء

0.52

OGND

0.51

 Мексичка

0.50

 kasarigan

0.50

 <<<<<<<<<<<<<<

0.50

 comuniques

0.50

Activations Density 0.003%

programming-related keywords and code structures

code-related identifiers

No Comments

No Known Activations