INDEX

Explanations

elements of mathematical notation and programming syntax

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

Loop indices (i, j, k) in code or math

np_acts-logits-general · gemini-2.0-flash

loop indices and delimiters

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

WriteTagHelper

-0.58

 ویکی‌پدی

-0.58

XmlAccessType

-0.56

acetic

-0.54

 Мексичка

-0.53

/**

-0.52

 мәкал

-0.51

 InputDecoration

-0.49

">//

-0.49

POSITIVE LOGITS

 annuel

0.46

 Haushalt

0.44

 terdekat

0.43

⅘

0.41

 fédé

0.41

 samfun

0.41

 sehari

0.41

MPN

0.40

paravant

0.40

 istrinya

0.40

Activations Density 1.511%

elements of mathematical notation and programming syntax

Loop indices (i, j, k) in code or math

loop indices and delimiters

No Comments

No Known Activations