INDEX

Explanations

attends to tokens related to specific attributes or functions from their corresponding generic terms

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.12

1:0.11

2:0.47

3:0.07

4:0.05

5:0.03

6:0.03

7:0.08

Negative Logits

Sucesor

-0.52

providedIn

-0.44

 المعيارى

-0.43

IntoConstraints

-0.40

iastes

-0.40

UserScript

-0.40

出版年

-0.38

 Réponses

-0.38

ValueStyle

-0.38

roek

-0.37

POSITIVE LOGITS

endsection

0.26

vea

0.25

 band

0.24

pravy

0.24

 intermédiaire

0.24

München

0.23

ابط

0.23

ิลปะ

0.23

 agricole

0.22

 Zeilen

0.22

Activations Density 1.102%

attends to tokens related to specific attributes or functions from their corresponding generic terms

No Comments

No Known Activations