INDEX

Explanations

attends to tokens that are URLs or code references from preceding tokens that are comments or text

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

0:0.16

1:0.14

2:0.18

3:0.04

4:0.04

5:0.02

6:0.11

7:0.28

Negative Logits

VIAF

-0.36

poptotic

-0.34

 CURIAM

-0.32

 ApiException

-0.31

 "..\..\

-0.30

MessageTagHelper

-0.29

Ecotoxicity

-0.29

unmodifiable

-0.29

Einzelnachweise

-0.29

NameInMap

-0.29

POSITIVE LOGITS

الإنجليزية

0.31

そ

0.27

 giuri

0.27

styleType

0.26

AlterField

0.26

IUrlHelper

0.25

sbereich

0.25

 Vors

0.24

 universel

0.24

 côtes

0.24

Activations Density 0.004%

attends to tokens that are URLs or code references from preceding tokens that are comments or text

No Comments

No Known Activations