INDEX

Explanations

terms associated with authority and noteworthy organizations

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

organizations and roles

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 nahilalakip

-0.84

CloseOperation

-0.79

 AssemblyCulture

-0.75

 <=",

-0.75

 يتيمه

-0.74

 Савезне

-0.74

Portail

-0.72

WithIOException

-0.72

LookAnd

-0.71

पया

-0.69

POSITIVE LOGITS

 broadest

0.34

par

0.33

bal

0.33

 ever

0.32

之一

0.32

 hardest

0.31

reg

0.29

its

0.29

 premier

0.29

Activations Density 0.016%

terms associated with authority and noteworthy organizations

organizations and roles

No Comments

No Known Activations