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

 ExecuteAsync

-0.77

 стаття

-0.50

SuppressLint

-0.44

__["

-0.41

 فريبيس

-0.39

 houſe

-0.39

 consultato

-0.38

 Werken

-0.37

HomeAsUpEnabled

-0.35

biť

-0.35

POSITIVE LOGITS

protoimpl

2.55

ValueStyle

0.73

 Bulletin

0.71

 ISNI

0.69

 EconPapers

0.68

verwijspagina

0.64



0.62

лтамалар

0.61

 kasarigan

0.60

bulletin

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.