INDEX

Explanations

document-related metadata and definitions

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

special characters and sequences

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

 تضيفلها

-0.50

 getItemId

-0.47

SharedDtor

-0.47

 surla

-0.46

 Elimin

-0.46

 deleteUser

-0.45

 iprot

-0.44

{{/

-0.43

createCanvas

-0.42

 endfor

-0.42

POSITIVE LOGITS

EndContext

0.48

 <<<<<<<<<<<<<<

0.45

tanleria

0.41

SourceChecksum

0.41

jde

0.38

 autorytatywna

0.38

INSEE

0.37

 SwitchCompat

0.37

ülé

0.37

\{\\

0.37

Activations Density 0.169%

document-related metadata and definitions

special characters and sequences

No Comments

No Known Activations