INDEX

Explanations

key programming concepts and structures related to functions, error handling, and system components

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

keyword followed by descriptor

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

oredCriteria

-0.77

ThroughAttribute

-0.62

exitRule

-0.60

 okuyayım

-0.49

MENAFN

-0.47

principalColumn

-0.45

์ตูน

-0.45

NPY

-0.44

 pageContext

-0.42

 protoimpl

-0.42

POSITIVE LOGITS

ícone

0.49

 χρή

0.47

شنایی

0.44

Vorlage

0.43

 Geruch

0.42

 selve

0.41

 kering

0.41

ilustración

0.41

fotografía

0.41

cartão

0.40

Activations Density 0.030%

key programming concepts and structures related to functions, error handling, and system components

keyword followed by descriptor

No Comments

No Known Activations