INDEX

Explanations

attends to auto-generated tokens from generated 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

0:0.13

1:0.20

2:0.14

3:0.19

4:0.12

5:0.02

6:0.09

7:0.08

Negative Logits

vivi

-0.30

ValueGenerated

-0.29

 NSCoder

-0.27

idiary

-0.27

swire

-0.25

 createSlice

-0.25

estinations

-0.25

 Parcelable

-0.24

 Fairly

-0.24

/***/

-0.24

POSITIVE LOGITS

PointerException

0.34

หวัด

0.31

SceneManagement

0.30

ंदीखरीदारी

0.30

 estimés

0.30

 Quora

0.29

dbo

0.27

ProgressHUD

0.27

олові

0.26

MethodManager

0.26

Activations Density 0.012%

attends to auto-generated tokens from generated tokens

No Comments

No Known Activations