INDEX

Explanations

specific codes, acronyms, or designations related to various subjects

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

acronyms or abbreviations

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

 katholischen

-0.40

iż

-0.38

 całym

-0.36

 which

-0.34

 constitucional

-0.34

zarchiwizowane

-0.33

 Bestimmung

-0.32

 Tinggi

-0.32

 Política

-0.32

 była

-0.32

POSITIVE LOGITS

 Roskov

0.86

 ویکی‌آمباردا

0.72

RTEE

0.71

AsUp

0.71

 noDo

0.71

 Paglinawan

0.69

 kasarigan

0.69

 kaarangay

0.68

astify

0.68

ьаж

0.66

Activations Density 0.064%

specific codes, acronyms, or designations related to various subjects

acronyms or abbreviations

No Comments

No Known Activations