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

FromArgb

-0.40

 Weih

-0.39

новниш

-0.37

encodeWith

-0.36

ội

-0.35

SharedDtor

-0.35

 goblins

-0.34

&__

-0.34

antMatchers

-0.34

 botanical

-0.33

POSITIVE LOGITS

lifecycle

0.64

IntoConstraints

0.62

Lifecycle

0.60

MutableLiveData

0.59

 Lifecycle

0.57

tagHelperRunner

0.57

 للاسماء

0.54

 createSlice

0.54

aarrggbb

0.54

 MutableLiveData

0.51

Activations Density 0.000%

No Known Activations

This feature has no known activations.