INDEX

Explanations

specific numerical identifiers associated with scholarly articles or studies

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

citation, author, year

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

enumi

-0.42

 تک

-0.36

บัติ

-0.36



-0.34

!*\

-0.34

 uiteindelijk

-0.34

 ""],

-0.33

rut

-0.33

 następ

-0.32

 δύ

-0.32

POSITIVE LOGITS

ArgsConstructor

0.78

 kasarigan

0.71

 EconPapers

0.65

contentLoaded

0.63

 ModelExpression

0.63

ViewImports

0.60

 initComponents

0.60

tonode

0.59

 Савезне

0.57

WarningLevel

0.56

Activations Density 0.028%

specific numerical identifiers associated with scholarly articles or studies

citation, author, year

No Comments

No Known Activations