INDEX

Explanations

formatting or structural elements in text documents

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

list items or new paragraphs

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

 geſch

-0.81

хьтан

-0.79

ſſung

-0.77

SharedDtor

-0.72

Хьажоргаш

-0.71



-0.71

 Italijani

-0.71

 Weiſe

-0.70

 ſehr

-0.69

səhifə

-0.69

POSITIVE LOGITS

 <<<<<<<<<<<<<<

0.43

 inoltre

0.33

 lisäksi

0.30

occasione

0.29

untung

0.28

Also

0.28

Therefore

0.28

Mask

0.27

Enlaces

0.27

Activations Density 0.049%

formatting or structural elements in text documents

list items or new paragraphs

No Comments

No Known Activations