INDEX

Explanations

numerical data associated with dates and chronological events

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

dates and years

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

AddTagHelper

-0.75

Hentet

-0.72

iſen

-0.70

Personensuche

-0.68

niſſe

-0.66

spesies

-0.66

$_(

-0.66

 ſind

-0.65

ſſung

-0.64

LookAnd

-0.63

POSITIVE LOGITS

 when

0.33

UnsafeEnabled

0.31

 following

0.31

 Gegenwart

0.28

 likewise

0.28

Lähteet

0.28

<code>

0.27

 said

0.27

Parcelize

0.26

 after

0.25

Activations Density 0.018%

numerical data associated with dates and chronological events

dates and years

No Comments

No Known Activations