INDEX

Explanations

references to recent periods or timeframes in research or studies

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

"recent" followed by a period of time

np_acts-logits-general · gemini-2.0-flash

recent followed by years, studies, or developments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 currently

-0.83

 later

-0.79

currently

-0.77

eventually

-0.76

 eventually

-0.75

 actuellement

-0.74

 presently

-0.74

 derzeit

-0.72

 später

-0.71

 attualmente

-0.68

POSITIVE LOGITS

 additions

0.86

MemoryWarning

0.86

 developments

0.81

 spate

0.80

 addition

0.80

 resourceCulture

0.78

 discoveries

0.77

 arrivals

0.75

 EconPapers

0.75

 trends

0.74

Activations Density 0.086%

references to recent periods or timeframes in research or studies

"recent" followed by a period of time

recent followed by years, studies, or developments

No Comments

No Known Activations