INDEX

Explanations

programming constructs related to user interface elements and their visibility states

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 geï

-0.37

 ansatte

-0.35

 voluto

-0.33

ļi

-0.31

 voorzien

-0.30

 curiosity

-0.30

 pertanian

-0.30

bardziej

-0.30

 curiosidad

-0.29

 инструмента

-0.29

POSITIVE LOGITS

 typelib

0.78

SBATCH

0.72

 NavController

0.63

argout

0.63

 onStop

0.63

 viewDidLoad

0.63

 onPause

0.62

 Biôgrafia

0.62

styleType

0.62

dafx

0.62

Activations Density 0.012%

programming constructs related to user interface elements and their visibility states

No Comments

No Known Activations