INDEX

Explanations

code comments and metadata in programming contexts

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

Code and user behavior at the start of a turn

np_acts-logits-general · gemini-2.0-flash

code inspections and suppressions

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

RenderAtEndOf

-0.62

EDEFAULT

-0.51

PerformLayout

-0.48

 Bioaccumulative

-0.48

 endwhile

-0.45

PhysRevD

-0.44

 ModelExpression

-0.44

 OFDb

-0.44

 UIApplication

-0.43

$_['

-0.42

POSITIVE LOGITS

 PhpStorm

0.71

'}';

0.54

cupertino

0.53

intellij

0.51

httphttps

0.47

"@/

0.42

setViewportView

0.42

})*/

0.40

>--}}

0.40

 Fieber

0.39

Activations Density 0.039%

code comments and metadata in programming contexts

Code and user behavior at the start of a turn

code inspections and suppressions

No Comments

No Known Activations