INDEX

Explanations

keywords and identifiers related to programming and software development

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

hardware/bus/asset keywords

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

TagMode

-0.73

 ویکی‌پدی

-0.69

 مرئيه

-0.60

 للمعارف

-0.59

UnusedPrivate

-0.59

writerow

-0.56

 otomatig

-0.55

ragalactic

-0.55

setVerticalGroup

-0.54

UrlResolution

-0.54

POSITIVE LOGITS

ellement

0.34

र्शन

0.33

 acert

0.32

azzle

0.32

ility

0.32

eker

0.31

 karş

0.31

を使用

0.30

ๆ

0.29

 gebruiken

0.28

Activations Density 0.090%

keywords and identifiers related to programming and software development

hardware/bus/asset keywords

No Comments

No Known Activations