INDEX

Explanations

references to Sundays and related time periods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 35-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_35/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.35.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

 Friday

-0.73

 Fridays

-0.69

Friday

-0.67

 friday

-0.65

friday

-0.65

 viernes

-0.62

 FRIDAY

-0.61

 vendredi

-0.57

 Thursday

-0.57

FRIDAY

-0.57

POSITIVE LOGITS

 Sunday

1.47

Sunday

1.41

sunday

1.26

 SUNDAY

1.25

SUNDAY

1.17

 sunday

1.13

 Monday

1.06

 zondag

1.03

Monday

1.03

 søndag

1.01

Activations Density 0.169%

references to Sundays and related time periods

No Comments

No Known Activations