INDEX

Explanations

frequency-related terms and phrases

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

"Every" followed by a time period

np_acts-logits-general · gemini-2.0-flash

recurring time periods

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

TintMode

-0.91

DockStyle

-0.70

Days

-0.67

RRect

-0.64

IVEREF

-0.61

 Days

-0.58

queryInterface

-0.58

msgSender

-0.57

Portail

-0.57

 francs

-0.56

POSITIVE LOGITS

routeProvider

0.61

 couple

0.60

 night

0.60

withIdentifier

0.55

 chance

0.55

few

0.55

войства

0.55

 time

0.54

 xong

0.53

nyt

0.53

Activations Density 0.106%

frequency-related terms and phrases

"Every" followed by a time period

recurring time periods

No Comments

No Known Activations