INDEX

Explanations

terms and phrases related to definitions and formal descriptions in a mathematical or theoretical context

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

definition

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

 gritos

-0.32

 joyas

-0.32

 uzyskać

-0.31

ClassNotFound

-0.31

 convain

-0.31

 Evaluación

-0.31

 zatr

-0.30

 onOptions

-0.30

Bromoform

-0.30

CSharp

-0.30

POSITIVE LOGITS

 definition

0.79

 Definition

0.77

Definition

0.75

DEFINITION

0.66

definition

0.66

XmlAccessType

0.66

IsMutable

0.66

 definitions

0.65

 definiert

0.64

 مشين

0.63

Activations Density 0.179%

terms and phrases related to definitions and formal descriptions in a mathematical or theoretical context

definition

No Comments

No Known Activations