INDEX

Explanations

mathematical symbols and terms

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

LaTeX commands and labels

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

++]=

-0.43

msgTypes

-0.40

heça

-0.40

strconv

-0.39

 endommag

-0.39

 zeros

-0.39

 MotionEvent

-0.38

ientôt

-0.38

pecabe

-0.38

 diputado

-0.37

POSITIVE LOGITS

 kasarigan

0.55

تقاوى

0.50

Personendaten

0.49

addPreferredGap

0.49

 насељу

0.49

 fubject

0.47

 تضيفلها

0.46

LayoutStyle

0.45

 cherchés

0.44

 AssemblyCompany

0.44

Activations Density 0.014%

mathematical symbols and terms

LaTeX commands and labels

No Comments

No Known Activations