INDEX

Explanations

attends to articles from corresponding tokens related to those articles

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.10

1:0.21

2:0.06

3:0.04

4:0.09

5:0.28

6:0.06

7:0.12

Negative Logits

msgTypes

-0.32

 ſmall

-0.31

 doubtnut

-0.31

Efq

-0.31

braith

-0.30

 houſe

-0.30

 itſelf

-0.30

 AttributeSet

-0.29

 uſed

-0.29

 purpoſe

-0.29

POSITIVE LOGITS

MLLoader

0.28

HtmlAttribute

0.26

interopRequire

0.25

 utafitiHapana

0.25

ContentAsync

0.25

setVerticalGroup

0.25

 الاطلاع

0.25

tiki

0.24

 knex

0.24

ScopeManager

0.24

Activations Density 0.042%

attends to articles from corresponding tokens related to those articles

No Comments

No Known Activations