INDEX

Explanations

programming-related terms and structures, specifically those associated with function definitions and applications in code

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

setup and initialization

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

fjspx

-0.48

 bambú

-0.44

offsetof

-0.42

OGND

-0.39

 jeugd

-0.39

 esclavos

-0.37

 montanha

-0.36

 actualidad

-0.36

 boneca

-0.35

 cuchara

-0.34

POSITIVE LOGITS

BeforeAll

0.85

AfterClass

0.81

BeforeClass

0.81

beforeAll

0.74

JvmStatic

0.58

 crê

0.53

ViewInit

0.51

 Reader

0.48

classmethod

0.47

JTable

0.47

Activations Density 0.001%

programming-related terms and structures, specifically those associated with function definitions and applications in code

setup and initialization

No Comments

No Known Activations