INDEX

Explanations

attends to tokens indicating contrast or distinction from subsequent corresponding tokens

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

0:0.10

1:0.13

2:0.10

3:0.05

4:0.04

5:0.02

6:0.07

7:0.46

Negative Logits

atoire

-0.27

soort

-0.26

 sucker

-0.25

pfung

-0.24

nerd

-0.23

 jadx

-0.23

 متعلقه

-0.23

udar

-0.23

tieri

-0.23

Activité

-0.22

POSITIVE LOGITS

contentLoaded

0.39

IUrlHelper

0.36

 otomatig

0.32

 Ostat

0.30

 فريبيس

0.29

 hänen

0.29

 particular

0.29

הערות

0.27

tedrale

0.27

webElement

0.27

Activations Density 0.269%

attends to tokens indicating contrast or distinction from subsequent corresponding tokens

No Comments

No Known Activations