INDEX

Explanations

mentions of legal or judicial processes and terminology

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

Numbers followed by roman numerals

np_acts-logits-general · gemini-2.0-flash

numbers and punctuation

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 surla

-0.99

 noDo

-0.89

 للمعارف

-0.85

出版年

-0.79

AndEndTag

-0.78

 transfieras

-0.78

MLLoader

-0.77

tagext

-0.77

MigrationBuilder

-0.76

verwijspagina

-0.75

POSITIVE LOGITS

index

0.33

 bağlantılar

0.33

0.28

 jagung

0.27

 czarna

0.27

today

0.27

 Stirn

0.26

 comigo

0.26

Activations Density 0.183%

mentions of legal or judicial processes and terminology

Numbers followed by roman numerals

numbers and punctuation

No Comments

No Known Activations