INDEX

Explanations

references to gender and associated roles or awards

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

or and other

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

CI

-0.40

 arro

-0.39

 يتيمه

-0.39

Traceback

-0.36

KE

-0.36

 kasarigan

-0.36

CI

-0.36

 arka

-0.35

Ris

-0.35

 kere

-0.35

POSITIVE LOGITS

 Wikimedijinoj

0.61

 Administrativna

0.56

 виправивши

0.54

ChildScrollView

0.53

 láser

0.50

 volcán

0.48

 ब्रेकडाउन

0.48

 ExecuteAsync

0.46

 virtuales

0.44

withIdentifier

0.44

Activations Density 0.121%

references to gender and associated roles or awards

or and other

No Comments

No Known Activations