INDEX

Explanations

attends to equation labels from earlier mathematical expressions

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

1:0.14

2:0.14

3:0.12

4:0.12

5:0.10

6:0.08

7:0.14

Negative Logits

 تضيفلها

-0.36

 mergeFrom

-0.31

setupUi

-0.31

AxisAlignment

-0.29

expandindo

-0.29

.~(\

-0.28

rdom

-0.28

 Childs

-0.28

 "..\..\..\

-0.28

 createState

-0.28

POSITIVE LOGITS

böz

0.31

ftrag

0.29

lím

0.27

ungguhnya

0.26

enfone

0.26

 Sita

0.26

enumi

0.25

 Wern

0.25

principalTable

0.25

ztály

0.25

Activations Density 0.376%

attends to equation labels from earlier mathematical expressions

No Comments

No Known Activations