INDEX

Explanations

attends to specific instances of the same lexical item from each token within contexts that include various technical or functional tokens

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

0:0.05

1:0.11

2:0.05

3:0.04

4:0.15

5:0.46

6:0.05

7:0.06

Negative Logits

DockStyle

-0.38

__(/*!

-0.35

chapper

-0.33

abstractmethod

-0.32

GIVEREF

-0.31

gulier

-0.30

endregion

-0.30

nonatomic

-0.28

RectangleBorder

-0.28

 للاسماء

-0.28

POSITIVE LOGITS

 Channing

0.40

 podjela

0.36

idemia

0.34

Климат

0.31

المناصب

0.31

 russes

0.30

évaluateur

0.29

 referenties

0.29

 Савезне

0.29

 Италијани

0.29

Activations Density 0.166%

attends to specific instances of the same lexical item from each token within contexts that include various technical or functional tokens

No Comments

No Known Activations