INDEX

Explanations

references to knowledge, learning, and scholarly roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atrième

-0.51

 reputations

-0.49

Tembelea

-0.48

 Picchu

-0.48

aibo

-0.48

 toppers

-0.47

 commentaries

-0.46

rbrakk

-0.45

 Traction

-0.45

 arşivlendi

-0.45

POSITIVE LOGITS

 himself

0.64

himself

0.57

 person

0.56

 Seorang

0.50

ceğine

0.50

 للمعارف

0.49

eseorang

0.49

osoba

0.47

 mwenye

0.47

 woman

0.46

Activations Density 0.705%

references to knowledge, learning, and scholarly roles

No Comments

No Known Activations