INDEX

Explanations

attends to reading-related tokens from temporal markers that give a specific date

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

0:0.07

1:0.10

2:0.10

3:0.11

4:0.10

5:0.05

6:0.26

7:0.17

Negative Logits

 CreateTagHelper

-0.40

########.

-0.36

homonymie

-0.35

 متعلقه

-0.34

laughs

-0.33

PYX

-0.32

ViewFeatures

-0.31

 незавершена

-0.31

drawSprites

-0.31

">'.$

-0.31

POSITIVE LOGITS

rrggbb

0.28

 GenerationType

0.27

 ingår

0.26

PullParser

0.25

 incluido

0.25

ẨM

0.25

Superview

0.24

naver

0.24

 caseros

0.24

simage

0.23

Activations Density 0.024%

attends to reading-related tokens from temporal markers that give a specific date

No Comments

No Known Activations