INDEX

Explanations

numerical data related to time attributes or identifiers

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

numbers ending in 5

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

 NSCoder

-0.57

存于互联网档案馆

-0.50

 ujednoznacz

-0.47

 EconPapers

-0.47

 Seeder

-0.46

الدراسه

-0.46

Personendaten

-0.45

 فريبيس

-0.43

 ennemi

-0.42

تقاوى

-0.41

POSITIVE LOGITS

FunctionFlags

0.49

GraphicsUnit

0.42

stateMutability

0.42

findpost

0.41

stran

0.39

inne

0.38

printStackTrace

0.38

AnchorStyles

0.38

lios

0.38

Phy

0.37

Activations Density 0.012%

numerical data related to time attributes or identifiers

numbers ending in 5

No Comments

No Known Activations