INDEX

Explanations

punctuation and formatting marks 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 @ 14-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_14/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.14.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 للمعارف

-0.54

Архівовано

-0.43

ISupport

-0.42

Билгалдахарш

-0.41

 Master

-0.40

 Helm

-0.40

Ito

-0.39

 General

-0.39

 Foreign

-0.38

 Suzuki

-0.37

POSITIVE LOGITS

 nemlig

0.59

 podró

0.52

 gebou

0.51

 ontdek

0.51

 asegur

0.50

 zicht

0.49

indeed

0.49

 tričko

0.48

—

0.48

 vertrou

0.48

Activations Density 0.278%

punctuation and formatting marks in text

No Known Activations