INDEX

Explanations

strings related to code documentation or declarations in programming languages

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

user

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

 تضيفلها

-0.59

 nahilalakip

-0.56

 незавершена

-0.54

 EnglishChoose

-0.52

UnusedPrivate

-0.52

 snippetHide

-0.51

ConfigureAwait

-0.50

PhysRev

-0.49

politan

-0.49

Enllaces

-0.49

POSITIVE LOGITS

__":

0.54

the

0.53

rungsseite

0.51

]")]

0.51

 buses

0.48

BibitemShut

0.48

</tfoot>

0.47

 wagons

0.47

})));

0.46

 myſelf

0.46

Activations Density 0.006%

strings related to code documentation or declarations in programming languages

user

No Comments

No Known Activations