INDEX

Explanations

terms related to data types and numerical precision

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

picked without replacement

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 OMITBAD

-0.55

 useAppContext

-0.52

parsedMessage

-0.47

ruptedException

-0.46

Життєпис

-0.42

yntaxException

-0.42

glMatrixMode

-0.42

UnusedPrivate

-0.41

кобрита

-0.40

 mergeFrom

-0.40

POSITIVE LOGITS

 autorytatywna

0.52

UserScript

0.46

 незавершена

0.46

nloa

0.45

fjspx

0.44

Vezi

0.43

 CanadaChoose

0.42

 kaarangay

0.42

Controllo

0.40

 pinulongan

0.39

Activations Density 0.022%

terms related to data types and numerical precision

picked without replacement

No Comments

No Known Activations