INDEX

Explanations

sections of code or programming language keywords

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

declarations or definitions within a programming context

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

card playing suits

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

érat

-0.50

 bringen

-0.47

 fidanz

-0.46

 puntata

-0.46

<>();

-0.44

 stop

-0.44

 parque

-0.44

Pant

-0.44

 Parks

-0.43

 ragazza

-0.43

POSITIVE LOGITS

InjectAttribute

0.67

GraphicsUnit

0.65

EndContext

0.65

findpost

0.60

sidemargin

0.60

 下午

0.59

cookieParser

0.57

ServletException

0.56

tableFuture

0.56

 Réponses

0.55

Activations Density 0.015%

sections of code or programming language keywords

declarations or definitions within a programming context

card playing suits

No Comments

No Known Activations