INDEX

Explanations

specific proper nouns and individuals mentioned within the text

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

names of people or variables

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

انجليز

-0.59

RenderAtEndOf

-0.57

 estekak

-0.56

DockStyle

-0.50

SequentialGroup

-0.49

ParallelGroup

-0.49

ChildScrollView

-0.48

الحياه

-0.47

Diwedd

-0.46

enumi

-0.45

POSITIVE LOGITS

 суп

0.43

 pères

0.42

merking

0.41

Anti

0.39

참고

0.39

 popoli

0.39

'{@

0.38

bukkit

0.38

PAYMENT

0.37

 Murch

0.37

Activations Density 0.053%

specific proper nouns and individuals mentioned within the text

names of people or variables

No Comments

No Known Activations