INDEX

Explanations

references to time, specifically instances indicating past events

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

"earlier"

np_acts-logits-general · gemini-2.0-flash

earlier this year/month/week/today

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 <>",

-0.63

opsida

-0.59

 beginnetje

-0.53

 sàng

-0.52

 gezicht

-0.49

 समीक्षाओं

-0.49

 opérés

-0.48

 становника

-0.48

Auszeichnungen

-0.47

 løpet

-0.45

POSITIVE LOGITS

MENAFN

0.70

CloseOperation

0.60

DoubleQuotes

0.59

 μή

0.58

Bronnen

0.58

 today

0.58

})()

0.58

 Roskov

0.57

InstanceState

0.55

Referències

0.55

Activations Density 0.077%

references to time, specifically instances indicating past events

"earlier"

earlier this year/month/week/today

No Comments

No Known Activations