INDEX

Explanations

specific types of characters or symbols, likely focusing on formatting and syntax in text

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

mathematical symbols 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-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

IntoConstraints

-0.34

NOR

-0.29

貝

-0.28

ɵɵ

-0.28

ativement

-0.28

BagLayout

-0.27

走

-0.27

YON

-0.26

BERS

-0.26

ULOS

-0.26

POSITIVE LOGITS

:✨

0.75

httphttps

0.70

 للاسماء

0.68

adaptiveStyles

0.67

Aholisi

0.62

 Meksiku

0.59

 betweenstory

0.59

 Photocase

0.58

 pinulongan

0.58

 gainera

0.57

Activations Density 0.001%

specific types of characters or symbols, likely focusing on formatting and syntax in text

mathematical symbols and punctuation

No Comments

No Known Activations