INDEX

Explanations

coding terminology and structure

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

categories or subcategories

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

 ModelExpression

-0.58

Gambas

-0.56

 الرياضيه

-0.55

 تضيفلها

-0.54

BeginContext

-0.53

enterOuterAlt

-0.52

GOTREF

-0.51

 cherchés

-0.50

AndEndTag

-0.49

Rhestr

-0.48

POSITIVE LOGITS

 autorytatywna

0.35

 kaarangay

0.35

نت

0.32

 autre

0.32

thè

0.30

ue

0.30

 ویکی‌پدیا

0.30

 rire

0.29

Mme

0.29

zupeł

0.29

Activations Density 0.191%

coding terminology and structure

categories or subcategories

No Comments

No Known Activations