INDEX

Explanations

code components and syntax elements related to programming and data structure management

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

code and math notation

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

şört

-0.69

 surla

-0.68

 оригіналу

-0.65

蚪

-0.60

ThroughAttribute

-0.59

LookAnd

-0.59

 deſſen

-0.58

adona

-0.57

<unused41>

-0.57

<unused79>

-0.57

POSITIVE LOGITS

 bougies

0.35

 supérieurs

0.34

↵↵

0.31

;;

0.30

 Palacio

0.30

insertBefore

0.29

 procès

0.29

 TestBed

0.29

 procé

0.29

ultados

0.28

Activations Density 0.000%

code components and syntax elements related to programming and data structure management

code and math notation

No Comments

No Known Activations