INDEX

Explanations

terms related to blinding methodologies and confidentiality in research studies

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

secure, confidential, or protected

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

RegressionTest

-0.64

 otomatig

-0.54

IContainer

-0.53

 utafitiHapana

-0.49

 négociations

-0.49

transQ

-0.48

 autorytatywna

-0.48

 useAppContext

-0.47

heça

-0.47

 peines

-0.47

POSITIVE LOGITS

EndInit

0.45

 Alembic

0.43

 AutoMapper

0.41

CppMethod

0.39

UnifiedTopology

0.39

IsUnicode

0.38

 unsigned

0.37

UnknownFields

0.37

MigrationBuilder

0.37

signed

0.37

Activations Density 0.064%

terms related to blinding methodologies and confidentiality in research studies

secure, confidential, or protected

No Comments

No Known Activations