INDEX

Explanations

references to scientific publications and citations

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

scientific journal names

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

Tikang

-0.58

DoubleQuotes

-0.56

 برانيه

-0.56

Clik

-0.54

الحياه

-0.53

出版年

-0.52

forChild

-0.51

quité

-0.51

 EconPapers

-0.50

XtraReports

-0.50

POSITIVE LOGITS

qrstuvwxyz

0.39

ьаж

0.38

Referències

0.38

 kyse

0.34

 companhia

0.32

ționale

0.32

 للاسماء

0.31

GEBURTS

0.30

 Năm

0.30

 autorytatywna

0.30

Activations Density 0.268%

references to scientific publications and citations

scientific journal names

No Comments

No Known Activations