INDEX

Explanations

phrases related to quality, usability, and performance

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

good quality and effectiveness

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

 مرئيه

-0.38

hasMoreElements

-0.37

Notae

-0.37

 utafitiHapana

-0.36

jar

-0.36

有什麼

-0.35

Personendaten

-0.35

fav

-0.34

tag

-0.33

有什么

-0.33

POSITIVE LOGITS

 OMITBAD

0.68

 reasonable

0.48

 raisonnable

0.48

tably

0.47

reasonable

0.47

ScopeManager

0.46

RenderAtEndOf

0.45

 Chwiliwch

0.44

Reasonable

0.43

 decent

0.43

Activations Density 0.077%

phrases related to quality, usability, and performance

good quality and effectiveness

No Comments

No Known Activations