INDEX

Explanations

attends to tokens related to "Windows" from tokens related to programming and development tools

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

0:0.09

1:0.11

2:0.13

3:0.08

4:0.08

5:0.03

6:0.11

7:0.34

Negative Logits

GEBURTSDATUM

-0.46

WebVitals

-0.36

UnifiedTopology

-0.33

errHandler

-0.31

 ModelExpression

-0.31

ագրություններ

-0.30

 يتيمه

-0.30

rungsseite

-0.30

 ferons

-0.29

 quelcon

-0.29

POSITIVE LOGITS

########.

0.24

 bodyParser

0.23

ZER

0.22

ESSOR

0.21

burberry

0.21

ANAS

0.21

MessageOf

0.20

 poda

0.20

amint

0.20

 inter

0.20

Activations Density 0.453%

attends to tokens related to "Windows" from tokens related to programming and development tools

No Comments

No Known Activations