INDEX

Explanations

terms related to specific design and functionalities of products

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

describes functionality

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

AnimationsModule

-0.45

 GenerationType

-0.42

addCriterion

-0.39

 entsch

-0.38

ausch

-0.38

Sto

-0.38

UseVisualStyle

-0.37

latego

-0.36

 consigo

-0.36

culata

-0.36

POSITIVE LOGITS

 autorytatywna

0.57

 EconPapers

0.50

 OFDb

0.48

 ivelany

0.48

 Савезне

0.47

 Италијани

0.47

dafx

0.45

참고

0.44

OGND

0.44

expandindo

0.43

Activations Density 0.092%

terms related to specific design and functionalities of products

describes functionality

No Comments

No Known Activations