INDEX

Explanations

terms related to various types of media and their attributes

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

media, events, or objects

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

 المعيارى

-0.68

 Púb

-0.48

GEBURTSDATUM

-0.46

XmlAccessType

-0.46

ciz

-0.46

Aerop

-0.45

 chré

-0.45

Wifi

-0.43

 جدًا

-0.42

Gps

-0.42

POSITIVE LOGITS

eens

0.58

Життєпис

0.54

 paramInt

0.54

sweise

0.51

iż

0.50

 گردد

0.50

ifty

0.48

Портал

0.47

还得

0.47

 calitate

0.47

Activations Density 0.165%

terms related to various types of media and their attributes

media, events, or objects

No Comments

No Known Activations