INDEX

Explanations

notations or labels related to authorship or publication in academic content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 14-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_14/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.14.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

 mijne

-0.50

 entendido

-0.48

 gekomen

-0.45

 européenne

-0.44

 sostenibilidad

-0.44

ņem

-0.43

 creatividad

-0.43

 descal

-0.43

 dança

-0.43

zelfde

-0.43

POSITIVE LOGITS

 Normdatei

0.66

 Administrativna

0.55

formik

0.55

 propOrder

0.54

 ब्रेकडाउन

0.50

CloseOperation

0.49

WireFormatLite

0.49

انيف

0.49

 protoimpl

0.49

reportWebVitals

0.48

Activations Density 0.180%

notations or labels related to authorship or publication in academic content

No Comments

No Known Activations