INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

 favor

-1.12

 neighborhood

-1.07

 neighborhoods

-1.05

Neighbors

-1.05

neighborhood

-1.04

favorite

-1.03

ighborhood

-1.03

 favorite

-1.02

 FAVOR

-1.01

favorable

-0.99

POSITIVE LOGITS

Whilst

1.45

 Whilst

1.35

 whilst

1.26

 organiser

1.18

 organisers

1.16

initialise

1.12

 realises

1.10

 organising

1.09

 realising

1.09

 organisation

1.08

Activations Density 0.000%

No Known Activations

This feature has no known activations.