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

 transport

-0.32

Dictionary

-0.29

 features

-0.29

 take

-0.29

 Form

-0.29

zeugen

-0.28

 sacrifice

-0.28

 calculation

-0.28

 Transport

-0.28

 Western

-0.28

POSITIVE LOGITS

#+#

7.88

:+:

2.28

+#+#

2.02

+#+

1.74

+:+

1.23

########.

1.20

 nahilalakip

1.16

 المعيارى

1.12

OGND

1.11

libft

1.10

Activations Density 0.000%

No Known Activations

This feature has no known activations.