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

 виправивши

-0.99

picasso

-0.58

 Alembic

-0.53

ValueGeneration

-0.44

Golf

-0.42

논

-0.41

 Golf

-0.35

">$

-0.35

<%@

-0.34

 golf

-0.32

POSITIVE LOGITS

Nationalité

3.53

 propOrder

2.91

HasIndex

1.98

SizeMode

1.16

HasForeignKey

1.10

aarrggbb

1.06

OGND

1.04

CloseOperation

0.99

 betweenstory

0.94

GEBURTSDATUM

0.93

Activations Density 0.000%

No Known Activations

This feature has no known activations.