INDEX

Explanations

attends to scientific and mathematical notation involving specific symbols or abbreviations from tokens representing explanations or descriptions of those symbols or topics

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

0:0.23

1:0.25

2:0.14

3:0.05

4:0.06

5:0.05

6:0.07

7:0.11

Negative Logits

 nakalista

-0.35

-0.29

 отношению

-0.28

NewUrlParser

-0.27

ografija

-0.27

-0.26

aarrggbb

-0.26

liyor

-0.25

ID

-0.25

POSITIVE LOGITS

󠁿

0.49

 Theſe

0.43

}';

0.40

 élas

0.40

?')

0.38

ſelves

0.38

ſelf

0.37

")));

0.37

 cdti

0.36

 viſ

0.36

Activations Density 0.384%

attends to scientific and mathematical notation involving specific symbols or abbreviations from tokens representing explanations or descriptions of those symbols or topics

No Comments

No Known Activations