INDEX

Explanations

programming and coding constructs related to data types and function parameters

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

length, size, and counts

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

SuppressMessage

-0.42

 podjela

-0.39

nativeElement

-0.39

ViewChild

-0.37

 يتيمه

-0.36

 Feldman

-0.36

MENAFN

-0.35

oredCriteria

-0.35

EndContext

-0.35

TabStop

-0.35

POSITIVE LOGITS

OGND

0.45

AddTagHelper

0.43

 avoient

0.43

 pouvoit

0.42

 abstrait

0.42

 auroit

0.42

DockStyle

0.42

 viewType

0.42

 vícti

0.41

 přek

0.41

Activations Density 0.022%

programming and coding constructs related to data types and function parameters

length, size, and counts

No Comments

No Known Activations