INDEX

Explanations

phrases related to numerical data and measurements in various contexts

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

non-english words

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

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

”

-0.40

”:

-0.39

nem

-0.37

adina

-0.36

%:

-0.36

/$',

-0.36

-0.34

]]:

-0.34

":

-0.34

 salt

-0.33

POSITIVE LOGITS

 myſelf

0.67

 indígen

0.62

 насељу

0.60

 auffi

0.60

Aholisi

0.58

 برانيه

0.58

ronpa

0.58

 auroit

0.58

 verſ

0.57

endphp

0.57

Activations Density 0.201%

phrases related to numerical data and measurements in various contexts

non-english words

No Comments

No Known Activations