INDEX

Explanations

attends to various tokens related to specific contextual markers from tokens that indicate structural or functional roles

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

0:0.51

1:0.14

2:0.07

3:0.05

4:0.07

5:0.03

6:0.04

7:0.06

Negative Logits

ControllerAdvice

-0.23

варь

-0.21

BuilderFactory

-0.21

ของคุณ

-0.20

 règne

-0.19

origen

-0.19

is

-0.19

 Cardona

-0.18

 besides

-0.18

 الأصل

-0.18

POSITIVE LOGITS

 pinulongan

0.36

IsMutable

0.34

 ComVisible

0.34

parsedMessage

0.32

MLLoader

0.32

хьтан

0.31

 Lightboxes

0.31

ServletException

0.31

 InputDecoration

0.31

)];

0.31

Activations Density 1.683%

attends to various tokens related to specific contextual markers from tokens that indicate structural or functional roles

No Comments

No Known Activations