INDEX

Explanations

attends to tokens related to specific terms from generalized elements in the sequences

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

0:0.45

1:0.15

2:0.12

3:0.04

4:0.06

5:0.03

6:0.03

7:0.07

Negative Logits

mybatisplus

-0.35

_;

-0.31

例文帳に追加

-0.30

fjspx

-0.30

@[

-0.27

}}_{

-0.26

 zase

-0.26

 ***!

-0.25

}}_

-0.25

rigada

-0.25

POSITIVE LOGITS

AutoresizingMask

0.41

verständlich

0.35

UnsafeEnabled

0.34

 MainAxisSize

0.33

">//

0.33

 doInBackground

0.33

alamus

0.33

TagMode

0.33

 ويكيپيديا

0.32

 disparu

0.32

Activations Density 0.433%

attends to tokens related to specific terms from generalized elements in the sequences

No Comments

No Known Activations