INDEX

Explanations

significant events or descriptors related to meetings, conferences, or formal gatherings

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

multilingual and technical jargon

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

PerformLayout

-0.57

 للمعارف

-0.51

-0.50

 EClass

-0.50

OGND

-0.49

RTSC

-0.49

RegressionTest

-0.49

typeorm

-0.48

fromnode

-0.47

POSITIVE LOGITS

怎么了

0.43

 culturelle

0.42

nehå

0.40

 compéti

0.40

pem

0.40

 extérieures

0.39

Pem

0.38

GIVEREF

0.38

 vectorielles

0.37

 coupable

0.37

Activations Density 0.245%

significant events or descriptors related to meetings, conferences, or formal gatherings

multilingual and technical jargon

No Comments

No Known Activations