INDEX

Explanations

methods and functions related to data handling and memory management

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

decode object, perform selector, open stream

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

Personendaten

-0.63

IntoConstraints

-0.59

 HasFactory

-0.50

 objectForKey

-0.50

 betweenstory

-0.48

 tableName

-0.45

؜

-0.45

IContainer

-0.43

Tikang

-0.43

InsertCommand

-0.43

POSITIVE LOGITS

ésult

0.54

 JADX

0.49

 Kuala

0.46

 stack

0.45

 تضيفلها

0.43

LLocation

0.43

ENEFITS

0.43

nœ

0.42

 fucking

0.42

 fuckin

0.42

Activations Density 0.041%

methods and functions related to data handling and memory management

decode object, perform selector, open stream

No Comments

No Known Activations