INDEX

Explanations

phrases related to academic or procedural language, particularly those used in research or documentation contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.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

 acestui

-0.45

exitRule

-0.44

THE

-0.44

 estekak

-0.44

 acestei

-0.41

jenigen

-0.40

 Grüsse

-0.39

 solches

-0.39

 này

-0.38

 desses

-0.38

POSITIVE LOGITS

the

3.16

 την

0.39

rethe

0.39

的

0.38

の

0.38

 את

0.38

const

0.37

의

0.35

 самого

0.35

 самой

0.35

Activations Density 8.615%

phrases related to academic or procedural language, particularly those used in research or documentation contexts

No Comments

No Known Activations