INDEX

Explanations

code definitions and function declarations

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

code definition and execution

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

ResumeLayout

-0.52

 Kräfte

-0.45

 alguno

-0.44

 المعيارى

-0.44

 natürliche

-0.40

 sánchez

-0.40

 Unterscheidung

-0.40

الدراسه

-0.40

 halben

-0.39

Portail

-0.39

POSITIVE LOGITS

 function

0.67

expandindo

0.57

function

0.52

ngdoc

0.50

PositiveButton

0.49

NgModule

0.48

InjectAttribute

0.48

BBM

0.47

OnClickListener

0.46

 Function

0.44

Activations Density 0.016%

code definitions and function declarations

code definition and execution

No Comments

No Known Activations