INDEX

Explanations

elements related to mathematical or statistical properties and results

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

numbers and punctuation

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

 pinulongan

-0.75

 Normdatei

-0.69

DockStyle

-0.63

sizeCache

-0.60

RegistryLite

-0.60

uxxxx

-0.59

 ProtoMessage

-0.57

richTextPanel

-0.56

RTLR

-0.56

 AssemblyCulture

-0.56

POSITIVE LOGITS

 tausend

0.31

 sospe

0.31

delà

0.31

diğimiz

0.29

 suspect

0.28

ợn

0.27

APT

0.27

futbol

0.27

 devriez

0.27

şek

0.26

Activations Density 0.580%

elements related to mathematical or statistical properties and results

numbers and punctuation

No Comments

No Known Activations