INDEX

Explanations

documents discussing recent publications or readings

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

 whatever

-0.31

 normal

-0.31

 thing

-0.30

}{*}{}

-0.28

 main

-0.28

 structure

-0.27

なり

-0.27

 organization

-0.27

 attitude

-0.26

 milj

-0.26

POSITIVE LOGITS

 للمعارف

0.75

 initComponents

0.68

Personensuche

0.67

ScopeManager

0.65

webElementGuid

0.63

〮

0.62

<unused68>

0.60

parsedMessage

0.60

<unused52>

0.60

<unused51>

0.60

Activations Density 0.981%

documents discussing recent publications or readings

No Comments

No Known Activations