INDEX

Explanations

specific measurements and adaptations related to different sizes or types of objects

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

line length, projectile, linear guide, size

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

 Capital

-0.31

 promise

-0.31

 entend

-0.30

 administration

-0.30

잖

-0.29

 Administration

-0.29

Capital

-0.29

 prov

-0.28

 recognition

-0.28

 administ

-0.28

POSITIVE LOGITS

 AssemblyCulture

0.81

 дописавши

0.74

Personensuche

0.69

 ExecuteAsync

0.68

principalColumn

0.66

 arşivlendi

0.65

#+#

0.65

 esternos

0.64

__*/

0.64

webElementGuid

0.64

Activations Density 0.064%

specific measurements and adaptations related to different sizes or types of objects

line length, projectile, linear guide, size

No Comments

No Known Activations