INDEX

Explanations

references to possessive pronouns and specific identifiers related to ownership in code contexts

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

my followed by a noun

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

makeConstraints

-0.56

ObjectContext

-0.54

OrNil

-0.54

principalTable

-0.54

SequentialGroup

-0.54

complexContent

-0.52

Tikang

-0.52

Datuak

-0.50

migrationBuilder

-0.50

featureID

-0.50

POSITIVE LOGITS

 cérami

0.45

My

0.41

My

0.39

 espirituales

0.39

your

0.38

Ljava

0.37

 personalizados

0.36

ๆ

0.36

Your

0.36

 homemade

0.36

Activations Density 0.013%

references to possessive pronouns and specific identifiers related to ownership in code contexts

my followed by a noun

No Comments

No Known Activations