INDEX

Explanations

phrases related to methods or processes of achieving something, particularly through usage or application

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

method or means

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

TabIndex

-0.33

GOTREF

-0.32

 diejenigen

-0.30

versions

-0.29

 versions

-0.28

 derfor

-0.27

 dagegen

-0.27

 løpet

-0.26

 Außerdem

-0.26

 außerdem

-0.26

POSITIVE LOGITS

хьтан

0.84

thâu

0.75

$_(

0.71

Vidite

0.69

 podjela

0.69

asteroide

0.67

ðsíða

0.66

AndroidJUnit

0.65

PasswordEncoder

0.65

 ſans

0.64

Activations Density 0.065%

phrases related to methods or processes of achieving something, particularly through usage or application

method or means

No Comments

No Known Activations