INDEX

Explanations

numerical data and indicators

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

roman numeral or numbered lists

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

 excepción

-0.29

 desgracia

-0.28

 zprá

-0.26

 exception

-0.26

løs

-0.24

 olvidar

-0.24

BeforeAll

-0.24

 juos

-0.24

 buds

-0.24

 concluye

-0.23

POSITIVE LOGITS

 Photocase

0.85

StructEnd

0.84

 للاسماء

0.81

 AssemblyCulture

0.79

 Normdatei

0.75

ьаж

0.73

 autorytatywna

0.72

 uLocal

0.70

Datuak

0.70

Activations Density 0.026%

numerical data and indicators

roman numeral or numbered lists

No Comments

No Known Activations