INDEX

Explanations

dates and numerical references in text

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RegressionTest

-0.75

 utafitiHapana

-0.69

queryInterface

-0.68

 PeEnEo

-0.67

 للمعارف

-0.67

 صوتيه

-0.65

DeleteBehavior

-0.65

ThroughAttribute

-0.64

Hentet

-0.63

Cyfeiriadau

-0.63

POSITIVE LOGITS

 Friday

0.42

May

0.41

March

0.40

May

0.40

Friday

0.40

 Sunday

0.38

June

0.37

Sunday

0.36

 March

0.36

 weekend

0.36

Activations Density 0.101%

dates and numerical references in text

No Comments

No Known Activations