INDEX

Explanations

phrases and words indicating the presence of a subject or referential expressions

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

more than

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

+#+

-0.61

 pinulongan

-0.59

 OFDb

-0.58

новништво

-0.57

expandindo

-0.57

 ujednoznacz

-0.55

Viitteet

-0.54

 Italijani

-0.53

 onCreateView

-0.51

XmlAccessType

-0.51

POSITIVE LOGITS

 ayrıca

0.42

 inoltre

0.39

また

0.37

 همچنین

0.36

zież

0.35

 Drapeau

0.35

 außerdem

0.35

 krim

0.34

medades

0.33

 также

0.33

Activations Density 0.038%

phrases and words indicating the presence of a subject or referential expressions

more than

No Comments

No Known Activations