INDEX

Explanations

attends to the term "original" from various references to authors, movies, algorithms, and studies

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

0:0.20

1:0.17

2:0.31

3:0.04

4:0.04

5:0.02

6:0.03

7:0.15

Negative Logits

Charsets

-0.30

 pubblici

-0.30

 שוליים

-0.29

sego

-0.28

BASEPATH

-0.28

 يتيمه

-0.28

 alimentaires

-0.27

 arşivlendi

-0.27

 supérieures

-0.27



-0.27

POSITIVE LOGITS

 ***!

0.32

]))

0.29

EndContext

0.27

navigationItem

0.27

])));

0.26

 foon

0.25

?>

0.24

()));

0.23

)))

0.23

 themſelves

0.22

Activations Density 0.193%

attends to the term "original" from various references to authors, movies, algorithms, and studies

No Comments

No Known Activations