INDEX

Explanations

references or reasons contributing to various phenomena or opinions

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

Follows "in part" or similar phrases

np_acts-logits-general · gemini-2.0-flash

part of

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AutoField

-0.56

 AssemblyCompany

-0.54

transQ

-0.52

󠁮

-0.51

jsxFileName

-0.50

 useRouter

-0.49

richTextPanel

-0.49

raulic

-0.49

UnknownFieldSet

-0.48

XmlAccessorType

-0.48

POSITIVE LOGITS

tamment

0.43

Kilder

0.39

Vanjske

0.37

 notamment

0.37

的重要

0.36

imageFile

0.35

Among

0.35

 egyik

0.35

VersionUID

0.34

ことも

0.34

Activations Density 0.367%

references or reasons contributing to various phenomena or opinions

Follows "in part" or similar phrases

part of

No Comments

No Known Activations