INDEX

Explanations

attends to contrasting traits from accompanying 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

0:0.17

1:0.17

2:0.27

3:0.09

4:0.06

5:0.02

6:0.04

7:0.13

Negative Logits

PerformLayout

-0.46

DockStyle

-0.46

 ویکی‌پدیا

-0.44

parsedMessage

-0.41

TypedDataSet

-0.41

 autorytatywna

-0.38

 CURIAM

-0.38

>{@

-0.37

MemoryWarning

-0.35

 protoimpl

-0.35

POSITIVE LOGITS

ių

0.24

 tästä

0.23

nių

0.23

bhan

0.22

 weiteren

0.21

URDAY

0.21

 asupra

0.21

ėmis

0.20

 weitere

0.20

irmar

0.19

Activations Density 0.378%

attends to contrasting traits from accompanying tokens

No Comments

No Known Activations