INDEX

Explanations

attends to tokens signifying possession or origin from related tokens denoting categories or types

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

0:0.11

1:0.15

2:0.37

3:0.07

4:0.06

5:0.02

6:0.06

7:0.10

Negative Logits

avigation

-0.35

IsContent

-0.33

 ویکی‌پدیا

-0.32

kháu

-0.31

ftagPool

-0.30

cyklopedia

-0.29

]:

-0.28

[]:

-0.28

jface

-0.28

</s>

-0.28

POSITIVE LOGITS

 esterni

0.29

richTextPanel

0.29

 perfons

0.26

 насељу

0.26

 purpoſe

0.26

WindowConstants

0.25

UnknownFields

0.25

axaca

0.24

 responder

0.24

Например

0.24

Activations Density 0.747%

attends to tokens signifying possession or origin from related tokens denoting categories or types

No Comments

No Known Activations