INDEX

Explanations

proper nouns, especially names and titles

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

surnames

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

PerformLayout

-0.52

Rhestr

-0.47

APOR

-0.45

 الاطلاع

-0.44

 noDo

-0.43

 podmín

-0.43

 oprot

-0.42

adaptiveStyles

-0.42

XmlType

-0.41



-0.41

POSITIVE LOGITS

 getSystem

0.42

ftagPool

0.39

 abstra

0.37

 memberId

0.37

 ThemeData

0.37

UserScript

0.35

 BaseActivity

0.35

WritableDatabase

0.35

ActivityResult

0.35

 parentId

0.34

Activations Density 0.029%

proper nouns, especially names and titles

surnames

No Comments

No Known Activations