INDEX

Explanations

attends to tokens that are numeric or mathematical expressions from tokens that represent variables or parameters

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

1:0.11

2:0.09

3:0.11

4:0.12

5:0.05

6:0.18

7:0.19

Negative Logits

ethene

-0.38

ACTERS

-0.31

########.

-0.30

liflower

-0.29

commutative

-0.28

 بيها

-0.28

ResumeLayout

-0.27

метров

-0.26

 Dorsal

-0.26

 snippetHide

-0.26

POSITIVE LOGITS

LookAnd

0.48

 мәкал

0.37

MLLoader

0.37

хьтан

0.36

)");

0.35

 OFDb

0.33

();)

0.33

")

0.32

 !”

0.32

contentLoaded

0.32

Activations Density 1.341%

attends to tokens that are numeric or mathematical expressions from tokens that represent variables or parameters

No Comments

No Known Activations