INDEX

Explanations

elements related to coding or programming instructions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

 kasarigan

-0.56

 newOwner

-0.51

orianCalendar

-0.47

 colectiva

-0.47

 Aufenthalt

-0.47

arriver

-0.47

 houſe

-0.47

 olev

-0.46

 indígen

-0.46

OOTDTY

-0.46

POSITIVE LOGITS

IMPORTED

0.42

zero

0.38

Дина

0.35

antd

0.35

0.34

 zero

0.33

 Mathieu

0.33

 Numerade

0.32

 determine

0.32

вропей

0.32

Activations Density 0.010%

elements related to coding or programming instructions

No Comments

No Known Activations