INDEX

Explanations

references to the word "tar" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

ruptedException

-0.50

CreateIndex

-0.47

:✨

-0.42

pendientes

-0.41

Viitteet

-0.41

Portail

-0.38

 ligiloj

-0.38

 ostavi

-0.38

IndentedString

-0.38

 otomatig

-0.37

POSITIVE LOGITS

tar

1.44

tar

1.32

Tar

1.30

Tar

1.29

TAR

1.08

TAR

1.05

 Taras

0.81

 Tariff

0.78

 tariff

0.77

Tara

0.77

Activations Density 0.412%

references to the word "tar" in various contexts

No Comments

No Known Activations