INDEX

Explanations

attends to features or attributes associated with tokens from corresponding contexts

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

0:0.14

1:0.15

2:0.13

3:0.07

4:0.04

5:0.02

6:0.07

7:0.35

Negative Logits

("-");

-0.28

WaitGroup

-0.27

 tủ

-0.27

ngths

-0.27

endregion

-0.27

 []);

-0.26

닛

-0.26

Geplaatst

-0.26

 bont

-0.25

quid

-0.25

POSITIVE LOGITS

$?

0.29

省市镇

0.28

 gawas

0.28

ToBounds

0.28

 doInBackground

0.27

?».

0.26

%?

0.26

 Winaray

0.26

 eternally

0.26

rmtree

0.26

Activations Density 0.553%

attends to features or attributes associated with tokens from corresponding contexts

No Comments

No Known Activations