INDEX

Explanations

attends to specific types of pointer or reference tokens from various struct or type definitions within code snippets

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.14

1:0.15

2:0.09

3:0.06

4:0.09

5:0.33

6:0.06

7:0.06

Negative Logits

즌

-0.33

cheibe

-0.32

Manifest

-0.32

BEM

-0.31

úrese

-0.31

bottomrule

-0.31

ˏ

-0.31

foregroundColor

-0.31

imed

-0.30

getColumnIndex

-0.30

POSITIVE LOGITS

 فريبيس

0.42

NOPQRST

0.39

 يتيمه

0.34

TagMode

0.31

 geslacht

0.30

 ویکی‌پدی

0.30

AndEndTag

0.29

EndGlobalSection

0.29

 Normdatei

0.29

 poveznice

0.28

Activations Density 0.102%

attends to specific types of pointer or reference tokens from various struct or type definitions within code snippets

No Comments

No Known Activations