INDEX

Explanations

technical terms and concepts related to programming and software development

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

region address cur

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

Negative Logits

 للمعارف

-0.55

bcryptjs

-0.51

zuführen

-0.45

 awakeFromNib

-0.44

 objectMapper

-0.44

thorax

-0.44

InstanceState

-0.43

󠁿

-0.43

 HasFactory

-0.42

 endforeach

-0.40

POSITIVE LOGITS

expandindo

0.58

HandleFunc

0.57

APIView

0.56

SharedDtor

0.54

principalColumn

0.50

 faldas

0.50

 Wicidata

0.49

WebServlet

0.49

 konus

0.47

 complejidad

0.45

Activations Density 0.004%

technical terms and concepts related to programming and software development

region address cur

No Known Activations