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

 noDo

-0.70

principalColumn

-0.66

 Infórmanos

-0.64

 témoig

-0.64

ⓧ

-0.63

IBOutlet

-0.63

 propOrder

-0.61

 ModelExpression

-0.60

webElementXpaths

-0.60

OGND

-0.60

POSITIVE LOGITS

1.00

 dorada

0.38

 najważ

0.38

URLException

0.35

httphttps

0.35

skosten

0.34

 parezca

0.34

 isKindOfClass

0.34

toprule

0.34

 "../../../

0.34

Activations Density 0.000%

No Known Activations

This feature has no known activations.