INDEX

Explanations

phrases indicating uncertainty or doubt

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Cyfarwyddwr

-0.47

AndEndTag

-0.43

 незавершена

-0.41

Kontrola

-0.40

ModelSerializer

-0.39

Rohy

-0.39

 écou

-0.38

Atsauces

-0.38

 actitud

-0.38

azgo

-0.36

POSITIVE LOGITS

 propOrder

0.52

(&:

0.52

]")]

0.47

 VIAF

0.46

ieteur

0.45

 nahilalakip

0.45

 connexes

0.41

IsOptional

0.40

ciaal

0.40

 proposes

0.38

Activations Density 0.546%

phrases indicating uncertainty or doubt

No Comments

No Known Activations