INDEX

Explanations

elements related to structured data, such as keys, values, and specific attributes in code

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

code blocks, IDs, and versions

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

 iprot

-0.65

 autorytatywna

-0.56

!*\

-0.55

ⓧ

-0.54

 ModelExpression

-0.54

onViewCreated

-0.54

 initComponents

-0.53

 Normdatei

-0.50

QMetaType

-0.49

 TestBed

-0.49

POSITIVE LOGITS

 Ambiental

0.40

 lisäksi

0.36

'{@

0.36

 rappeler

0.35

 vastaan

0.34

 cómodo

0.34

 oignon

0.33

 découvertes

0.32

 abuelos

0.32

 löytyy

0.31

Activations Density 0.113%

elements related to structured data, such as keys, values, and specific attributes in code

code blocks, IDs, and versions

No Comments

No Known Activations