INDEX

Explanations

attends to nouns related to locations or structures from specific tokens that appear later in the text

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

1:0.09

2:0.10

3:0.45

4:0.07

5:0.03

6:0.06

7:0.07

Negative Logits

PreferredItem

-0.44

AsUp

-0.34

 InputDecoration

-0.33

ásban

-0.32

 الدولى

-0.32

enschappelijke

-0.30

Javadoc

-0.29

BagConstraints

-0.29

PropertyChanging

-0.29

 externi

-0.29

POSITIVE LOGITS

IGraphics

0.47

NewGuid

0.42

__*/

0.40

éraire

0.40

 Collegiate

0.39

 مشين

0.39

jsxFileName

0.38

vedì

0.38

ReusableCell

0.37

NOPQRST

0.37

Activations Density 0.838%

attends to nouns related to locations or structures from specific tokens that appear later in the text

No Comments

No Known Activations