INDEX

Explanations

occurrences of the letter 's' in various contexts

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

names followed by roles

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

IntoConstraints

-0.67

RegressionTest

-0.62

 ujednoznacz

-0.58

GOTREF

-0.57

 AssemblyCulture

-0.56

jspx

-0.56

 StatelessWidget

-0.54

 ContentValues

-0.53

internalType

-0.51

MessageOf

-0.50

POSITIVE LOGITS

 alumni

0.45

รุ่น

0.41

 navideña

0.38

wakili

0.36

 Alumni

0.35

 pimpinan

0.35

 prise

0.35

 chief

0.35

 correspondents

0.34

 president

0.34

Activations Density 0.048%

occurrences of the letter 's' in various contexts

names followed by roles

No Comments

No Known Activations