INDEX

Explanations

attends to general features or descriptors marked by specific pronouns or demonstratives from detailed or specific 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

2:0.39

3:0.07

4:0.05

5:0.03

6:0.04

7:0.10

Negative Logits

 ProtoMessage

-0.37

InjectAttribute

-0.37

 مشين

-0.32

 nemico

-0.32

اریخ

-0.31

BagLayout

-0.31

 ویکی‌پدیا

-0.30

 bonté

-0.30

 rêver

-0.29

 intptr

-0.29

POSITIVE LOGITS

inander

0.38

wamy

0.29

CGRectMake

0.28

ítimo

0.28

HasAnnotation

0.28

 SWIFT

0.28

zewod

0.27

vieja

0.27

وعة

0.27

 kroner

0.27

Activations Density 0.544%

attends to general features or descriptors marked by specific pronouns or demonstratives from detailed or specific contexts

No Comments

No Known Activations