INDEX

Explanations

proper nouns, particularly names of people, organizations, and titles

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

Preceding a phrase with a definition/description

np_acts-logits-general · gemini-2.0-flash

forms of to be verbs

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.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

出版年

-0.58

 kaarangay

-0.58

 CreateTagHelper

-0.54

NameInMap

-0.53

GOTREF

-0.53

rungsseite

-0.52

 AssemblyCulture

-0.51

CopyWith

-0.51

ingeki

-0.49

цездатний

-0.47

POSITIVE LOGITS

is

1.42

 adalah

1.09

 merupakan

1.07

 является

1.02

 represents

0.85

là

0.82

 consists

0.82

 ialah

0.81

was

0.80

是一个

0.80

Activations Density 1.029%

proper nouns, particularly names of people, organizations, and titles

Preceding a phrase with a definition/description

forms of to be verbs

No Comments

No Known Activations