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

 صوتيه

-1.08

 autorytatywna

-0.99

OGND

-0.91

MLLoader

-0.90

########.

-0.85

 otomatig

-0.82

AccessorTable

-0.79

ロウィン

-0.79

 imagui

-0.79

 CreateTagHelper

-0.78

POSITIVE LOGITS

 another

0.44

 Svensson

0.36

 Erhöhung

0.33

 Befehl

0.33

 also

0.33

another

0.33

0.32

Another

0.32

 Salud

0.32

 Verlust

0.32

Activations Density 0.000%

No Known Activations

This feature has no known activations.