INDEX

Explanations

technical terms related to programming and custom implementations

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

implementing from scratch

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

cry

-0.37

Cry

-0.37

 feature

-0.34

cry

-0.34

PickerController

-0.32

 ratio

-0.31

 inoc

-0.30

ReusableCell

-0.30

tshire

-0.30

 emend

-0.29

POSITIVE LOGITS

 незавершена

0.73

awtextra

0.67

tagHelperRunner

0.59

 ProtoMessage

0.58

adaptiveStyles

0.56

 Réponses

0.56

IContainer

0.52

 étoient

0.52

BagLayout

0.52

Carriera

0.51

Activations Density 0.187%

technical terms related to programming and custom implementations

implementing from scratch

No Comments

No Known Activations