INDEX

Explanations

attends to tokens marked with certain key patterns, specifically focusing on those related to functions or programming concepts from corresponding path 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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.09

1:0.11

2:0.11

3:0.13

4:0.10

5:0.08

6:0.18

7:0.15

Negative Logits

AxisAlignment

-0.27

Portale

-0.26

帖最后由

-0.26

]")]

-0.25

})$}

-0.25

็บ

-0.25

yorsunuz

-0.25

脚注の使い方

-0.24

Extinguishing

-0.24

ollection

-0.23

POSITIVE LOGITS

Personensuche

0.37

Kariera

0.27

yntaxException

0.26

NewRow

0.26

createQuery

0.26

laka

0.25

 fune

0.25

uoš

0.24

 cazul

0.24

isRequired

0.24

Activations Density 0.038%

attends to tokens marked with certain key patterns, specifically focusing on those related to functions or programming concepts from corresponding path tokens

No Comments

No Known Activations