INDEX

Explanations

references to historical or chronological context

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

TagMode

-0.74

AddTagHelper

-0.73

 HttpNotFound

-0.68

ünst

-0.68

 estekak

-0.63

#+#

-0.63

شهاد

-0.62

complexContent

-0.57

providedIn

-0.57

<?

-0.57

POSITIVE LOGITS

 WWII

1.23

 colonial

1.13

 medieval

1.09

 Medieval

1.02

 ancient

1.01

 nineteenth

1.00

WWI

1.00

 Victorian

0.99

 Nazi

0.94

 Colonial

0.94

Activations Density 0.724%

references to historical or chronological context

No Comments

No Known Activations