INDEX

Explanations

references to numbered lists and mathematical expressions

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

numbers and roman numerals

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

-0.63

OGND

-0.60

 незавершена

-0.58

 мәкал

-0.58

Derbyniad

-0.57

lapsingToolbar

-0.56

 gyhoeddwyd

-0.52

 CURIAM

-0.51

 onBind

-0.51

 sumpay

-0.51

POSITIVE LOGITS

 <<<<<<<<<<<<<<

0.37

[*]

0.36

 للاسماء

0.32

 proposition

0.32

version

0.31

ут

0.31

 Thirdly

0.30

 papan

0.30

konzept

0.30

Twice

0.29

Activations Density 0.102%

references to numbered lists and mathematical expressions

numbers and roman numerals

No Comments

No Known Activations