INDEX

Explanations

attends to menu-related tokens from various programming context tokens

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

1:0.11

2:0.06

3:0.06

4:0.06

5:0.07

6:0.07

7:0.43

Negative Logits

 >=",

-0.36

MigrationBuilder

-0.33

 beginnetje

-0.30

CloseOperation

-0.29

 transférez

-0.28

yl

-0.27

PyTuple

-0.27



-0.27

ItemBackground

-0.26

 تضيفلها

-0.26

POSITIVE LOGITS

})]

0.33

 ouvriers

0.31

 quelquefois

0.29

 Jove

0.29

□

0.29

 couverte

0.28

("")]

0.28

SRT

0.28

 aussieht

0.28

 resourceCulture

0.27

Activations Density 0.001%

attends to menu-related tokens from various programming context tokens

No Comments

No Known Activations