INDEX

Explanations

technical terms and concepts related to architecture and processing in systems

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

processed, studied, architecture, investigated

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

AndEndTag

-0.69

 CreateTagHelper

-0.60

 ExecuteAsync

-0.55

 Superhost

-0.54

ReusableCell

-0.52

annique

-0.51

IsMutable

-0.49

Personendaten

-0.49

transQ

-0.48

 unknownFields

-0.48

POSITIVE LOGITS

 agreed

0.36

queryInterface

0.33

arlı

0.33

 quedado

0.32

Zam

0.31

 Sardegna

0.31

querySelectorAll

0.31

 honn

0.30

webElementXpaths

0.30

WaitGroup

0.30

Activations Density 0.471%

technical terms and concepts related to architecture and processing in systems

processed, studied, architecture, investigated

No Comments

No Known Activations