INDEX

Explanations

phrases involving conditional or contextual relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 الرياضيه

-0.79

脚注の使い方

-0.72

setVerticalGroup

-0.71



-0.71

PerformLayout

-0.67

Parcelize

-0.67

Cubit

-0.67

BeginInit

-0.64

 StatelessWidget

-0.64

 externi

-0.64

POSITIVE LOGITS

Hentet

0.63

 whom

0.53

 Besoin

0.49

 whose

0.48

who

0.46

[][

0.46

 which

0.45

cui

0.44

venidos

0.43

whose

0.42

Activations Density 0.428%

phrases involving conditional or contextual relationships

No Comments

No Known Activations