INDEX

Explanations

attends to tokens starting with "She" from corresponding tokens that follow after the activation

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

0:0.08

1:0.01

2:0.05

3:0.03

4:0.58

5:0.13

6:0.03

7:0.05

Negative Logits

 nakalista

-0.29

 kona

-0.28

ровано

-0.28

 MongoClient

-0.27

 Grand

-0.27

Advertisement

-0.27

isini

-0.27

grand

-0.26

 wife

-0.26

 Armour

-0.26

POSITIVE LOGITS

تقاوى

0.38

 AttributeSet

0.36

GraphicsUnit

0.35

 EnglishChoose

0.34

AtIndexPath

0.34

BitmapFactory

0.33

Vidite

0.33

 gezet

0.33

MockMvc

0.33

ChildScrollView

0.33

Activations Density 0.138%

attends to tokens starting with "She" from corresponding tokens that follow after the activation

No Comments

No Known Activations