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

Przeczytaj

-0.43

しの

-0.43

 dials

-0.41

 ionizing

-0.40

 <<<<<<<<<<<<<<

-0.40

 mitge

-0.39

 fört

-0.38

libft

-0.38

PositiveButton

-0.37

 宝宝

-0.36

POSITIVE LOGITS

Mvc

2.47

MVC

0.97

MVC

0.95

mvc

0.93

mvc

0.88

IActionResult

0.68

StoryboardSegue

0.68

 OFDb

0.63

 AspNetCore

0.62

 ActionResult

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.