INDEX

Explanations

JavaScript-related keywords and constructs, particularly for function definitions and AJAX requests

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

web application code, JavaScript

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

 nakalista

-0.76

ंदीखरीदारी

-0.60

AnchorStyles

-0.57

 GenerationType

-0.55

 autorytatywna

-0.54

 EconPapers

-0.54

GEBURTSDATUM

-0.51

 resourceCulture

-0.51

LayoutStyle

-0.50

 дописавши

-0.50

POSITIVE LOGITS

 learned

0.39

 chicos

0.38

 τραγ

0.35

 connaissances

0.34

 ourselves

0.33

 niñas

0.33

 myself

0.33

Становништво

0.33

 itself

0.32

 chicas

0.32

Activations Density 0.066%

JavaScript-related keywords and constructs, particularly for function definitions and AJAX requests

web application code, JavaScript

No Comments

No Known Activations