INDEX

Explanations

phrases indicating reasoning or conclusion

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

preceding verbs of opinion or conclusion

np_acts-logits-general · gemini-2.0-flash

to claim or conclude

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

المناصب

-0.39

done

-0.34

 East

-0.32

 Autorizaciones

-0.31

 Stillwater

-0.31

lives

-0.30

بوابة

-0.30

EH

-0.30

he

-0.29

FU

-0.29

POSITIVE LOGITS

HtmlAttribute

0.63

 defaultstate

0.63

CppCodeGen

0.60

SharedDtor

0.58

 noted

0.57

 nhắc

0.56

 insinu

0.56

 obſ

0.55

 suggest

0.54

 noting

0.54

Activations Density 0.781%

phrases indicating reasoning or conclusion

preceding verbs of opinion or conclusion

to claim or conclude

No Comments

No Known Activations