INDEX

Explanations

evidence of numerical or coded representations, likely related to data formats or programming

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

screen region codes

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

SourceChecksum

-0.66

ⓧ

-0.65

InputBorder

-0.61

]='\

-0.57

MessageTagHelper

-0.56

AndEndTag

-0.56

كويكب

-0.55

 CanadaChoose

-0.55

yntaxException

-0.53

 surla

-0.53

POSITIVE LOGITS

SuppressMessage

0.98

DoNot

0.38

 schermata

0.34

Constructors

0.34

 bolsillo

0.34

 screen

0.33

 brush

0.32

Shades

0.31

 ocupado

0.31

 coscienza

0.31

Activations Density 0.001%

evidence of numerical or coded representations, likely related to data formats or programming

screen region codes

No Comments

No Known Activations