INDEX

Explanations

structured data or formatted entries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 estekak

-0.62

IUrlHelper

-0.59

awaiter

-0.49

Filmographie

-0.47

providedIn

-0.46

 Chwiliwch

-0.44

 <>",

-0.44

ChildScrollView

-0.44

 ویکی‌پدی

-0.43

 juridiques

-0.43

POSITIVE LOGITS

UrlResolution

0.46

 MonoBehaviour

0.46

Radi

0.45

 Baş

0.41

 scra

0.41

kháu

0.40

Jud

0.40

 radi

0.40

叶修

0.40

BlockPos

0.39

Activations Density 3.263%

structured data or formatted entries

No Comments

No Known Activations