INDEX

Explanations

phrases indicating advice or recommendations regarding behavior and etiquette

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

not a, not confusing, when proposing, very confusing, many companies, often called

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

expandindo

-0.56

MessageOf

-0.42

 odkazy

-0.33

RenderAtEndOf

-0.31

 ParseException

-0.31

NVIC

-0.30

 Broadcast

-0.30

autoreleasepool

-0.29

 suplemento

-0.29

poj

-0.29

POSITIVE LOGITS

GraphicsUnit

0.80

 kasarigan

0.70

 Italijani

0.56

Personensuche

0.55

InitVars

0.49

Referencies

0.49

 dår

0.48

síða

0.48

gæ

0.48

 EconPapers

0.48

Activations Density 0.023%

phrases indicating advice or recommendations regarding behavior and etiquette

not a, not confusing, when proposing, very confusing, many companies, often called

No Comments

No Known Activations