INDEX

Explanations

identifiers and data structures typically found in programming or coding contexts

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

 ✭✭

-0.68

 ***!

-0.61

 Савезне

-0.54

 Infórmanos

-0.54

 wireType

-0.53

(!__

-0.52

 Senhora

-0.52

存于互联网档案馆

-0.51

 виправивши

-0.50

 bezeichneter

-0.50

POSITIVE LOGITS

RefNanny

0.38

SourceChecksum

0.38

ruvate

0.32

 NgModule

0.30

 davon

0.30

 CGPoint

0.30

DockStyle

0.28

จับ

0.28

rieks

0.28

 léka

0.28

Activations Density 0.296%

identifiers and data structures typically found in programming or coding contexts

No Comments

No Known Activations