INDEX

Explanations

elements related to database actions and attributes

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

efois

-0.52

 communiquez

-0.49

oubliez

-0.46

 prawda

-0.46

ReusableCell

-0.45

étoit

-0.45

 Grüßen

-0.43

étaient

-0.43

 مرئيه

-0.42

خدا

-0.42

POSITIVE LOGITS

0.62

AutoScaleMode

0.58

А

0.57

AnchorTagHelper

0.56

 Aces

0.51

addHeader

0.50

AEA

0.49

 AppBar

0.49

Aos

0.49

Activations Density 3.246%

elements related to database actions and attributes

No Comments

No Known Activations