INDEX

Explanations

key technical terms and specific identifiers in programming contexts

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

execute out

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

yntaxException

-0.46

WireFormatLite

-0.44

hamilan

-0.43

 circonstances

-0.43

 FetchType

-0.40

thansa

-0.40

 entera

-0.37

 ""],

-0.37

pośred

-0.36

 inteligen

-0.36

POSITIVE LOGITS

Hentet

0.53

webElementXpaths

0.46

 للمعارف

0.44

ValueStyle

0.44

sizeCache

0.43

 صوتيه

0.43

RegistryLite

0.43

:✨

0.43

章节错误

0.43

 jspb

0.41

Activations Density 0.034%

key technical terms and specific identifiers in programming contexts

exec**ute out**

No Comments

No Known Activations

execute out