INDEX

Explanations

punctuation marks and formatting symbols

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

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

 Administrativna

-0.60

 Италијани

-0.58

 kaarangay

-0.54

Panamoan

-0.54

andoah

-0.53

يكب

-0.52

ilares

-0.51

 noDo

-0.51

intios

-0.50

expandindo

-0.50

POSITIVE LOGITS

 informes

0.40

matchCondition

0.39

+#+

0.38

 Staff

0.37

Tre

0.36

 Suit

0.36

chaleco

0.36

 plomo

0.36

 yani

0.35

jsxFileName

0.35

Activations Density 0.613%

punctuation marks and formatting symbols

symbols and punctuation

No Comments

No Known Activations