INDEX

Explanations

attends to the token "CALIBRATION" from various tokens referring to specific points or states

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

0:0.79

1:0.00

2:0.01

3:0.01

4:0.09

5:0.02

6:0.01

7:0.02

Negative Logits

 المعيارى

-0.37

 JpaRepository

-0.35

 kasarigan

-0.35

tvguidetime

-0.35

 الرياضيه

-0.33

 nakalista

-0.32

ագրություններ

-0.31

 BoxFit

-0.30

 povezave

-0.30

 betweenstory

-0.30

POSITIVE LOGITS

rawDesc

0.29

KEYCODE

0.27

endregion

0.24

乎

0.24

PYX

0.24

MutableLiveData

0.23

<![

0.23

colgroup

0.23

<thead>

0.23

Asp

0.22

Activations Density 6.732%

attends to the token "CALIBRATION" from various tokens referring to specific points or states

No Comments

No Known Activations