INDEX

Explanations

names or proper nouns, particularly those related to individuals

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

names of people

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

 PropTypes

-0.44

ſelves

-0.42

يميديا

-0.41

 itſelf

-0.41

 makinesi

-0.40

 CanadaChoose

-0.40

 Knochen

-0.40

Tikang

-0.40

 sèche

-0.40

кобрита

-0.40

POSITIVE LOGITS

queryInterface

0.53

MigrationBuilder

0.45

InjectAttribute

0.44

 betweenstory

0.43

knex

0.43

 Autorizaciones

0.42

RTSC

0.41

ArgsConstructor

0.40

NameInMap

0.40

InsertCommand

0.40

Activations Density 0.038%

names or proper nouns, particularly those related to individuals

names of people

No Comments

No Known Activations