INDEX

Explanations

certain HTML or programming syntax elements

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

user turn indicator

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

-0.55

 AssemblyCompany

-0.52

findpost

-0.50

AntiForgeryToken

-0.50

CORBA

-0.49

MessageTagHelper

-0.49

queryInterface

-0.48

WithIOException

-0.47

 AssemblyTitle

-0.46

endregion

-0.46

POSITIVE LOGITS

 kasarigan

0.52

Kaynakça

0.48

 Stimmung

0.46

 vectorielles

0.45

ResourceManager

0.45

 navideños

0.44

 judiciales

0.43

Fordítás

0.41

ResourceId

0.41

 espirituales

0.40

Activations Density 0.000%

certain HTML or programming syntax elements

user turn indicator

No Comments

No Known Activations