INDEX

Explanations

No Explanations Found

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

 nonUne

-0.35

 terá

-0.32

“

-0.30

 kolei

-0.30

|}{}

-0.28

 besonderer

-0.28

 élus

-0.28

 tačiau

-0.28

 regeringen

-0.28

፩

-0.28

POSITIVE LOGITS

 незавершена

0.84

 للمعارف

0.75

 Chwiliwch

0.67

 newOwner

0.66

IUrlHelper

0.65

ValueStyle

0.64

WithIOException

0.63

:✨

0.61

BufferException

0.60

BeginInit

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.