INDEX

Explanations

phrases related to documentation and data organization

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

related to or following conjunctions

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

Ἁ

-0.45

المناصب

-0.40

pherals

-0.40

spl

-0.40

Aggregate

-0.38

Autoritní

-0.38

umab

-0.36

ospor

-0.36

 unbekannt

-0.36

?>

-0.35

POSITIVE LOGITS

 surla

0.54

Хьажоргаш

0.50

AndEndTag

0.46

 informée

0.45

 abiertos

0.44

enity

0.43

 مرئيه

0.42

rungsseite

0.41

scriptcase

0.39

 informally

0.38

Activations Density 0.062%

phrases related to documentation and data organization

related to or following conjunctions

No Comments

No Known Activations