Dimiter Milushev

[Dimiter Milushev @haskool]

5/Everything is open source:

Dataset: https://t.co/YqYKuMBVH8 Model: https://t.co/Z5xfdDj0eE Code: https://t.co/tOAYBVEomL Writeup: https://t.co/M4Ot1yZN4c

[Dimiter Milushev @haskool]

1/I built an open-source prompt injection detector. DeBERTa-v3-base trained on a curated dataset of 54K samples.
96.1% detection rate at 0.1% false positive rate — same architecture as ProtectAI v2, but +17pp from better training data.

[Dimiter Milushev @haskool]

4/It's not a silver bullet. Jailbreak detection is the weakest at 93.5%. English-dominant. Single-turn only. And at real-world attack rates (<0.1%), even 0.1% FPR means false alarms outnumber real detections.
It's Layer 1 in a defense stack — fast pre-filter, not the final word.

[Dimiter Milushev @haskool]

@capitalcom . Ä

[Mark Boss @markb_boss]

NeRD and Neural-PIL enable decompositions into shape, BRDF, and lighting. However, they require camera poses. Common in-the-wild collections are not only in multiple illuminations but also locations. Here, COLMAP fails. SAMURAI aims to fix this: https://t.co/gAi1dZBs8M

🧵

[Vesselin Dremdjiev @vdremdjiev]

🔴 Светът има проблем - руският паранаичният диктатор #Путин започна война срещу #Украйна. Заплашва и всички по #света, които му се противопоставят.
Не, не е сценарий на евтин екшън.

[Dimiter Milushev @haskool]

@jeremyphoward I am interested in translating the next one i Bulgarian.

[Jack Crook @jackcr]

My non-inclusive tweet for the year: I'm looking forward to the time when all of the new year's resolution gym members realize a certain amount of commitment is needed and most stop going. It's far too crowded during the first few weeks of a new year.

[Dimiter Milushev @haskool]

https://t.co/MzwkQMTNWO

[Dimiter Milushev @haskool]

No substitute to expert knowledge
... https://t.co/FZgJ6qBhyU

[Thang Luong @lmthang]

Excited to share our #emnlp2018 work: SOTA on many NLP tasks where performances already feel saturated! ideas: models self-train on unlabeled data by refraining access from full representations of text sequences + multitask-traing. by @clark_kev, @chrmanning, @quocleix & I.

[Dimiter Milushev @haskool]

@BSidesZurich Many thanks for organizing, I had a great time at my first BSides Zurich. Great job on selecting a really nice mix of interesting speakers who were able to cover a wide spectrum of topics!

[Dimiter Milushev @haskool]

Really cool https://t.co/dfVbnpCP6L

[Sebastian Ruder @seb_ruder]

Transfer learning with language models is getting hot! 🔥New state-of-the-art results today by two different research groups: Trinh and Le (Google) on the Winograd challenge and Radford et al. (OpenAI) on a diverse range of tasks.
https://t.co/UKwmyslQbf
https://t.co/Xhh01WZkzy

[Hugh Ross @RTB_HRoss]

NASA just released these 8 images of Jupiter taken by the Juno spacecraft, the highest resolution, most color detailed to date.

[Sebastian Ruder @seb_ruder]

"I interviewed at 5 top companies in 5 days, and got 5 job offers" - honest article about interviews & preparation https://t.co/kzKbxE9zH0

[Dimiter Milushev @haskool]

Really nice talk! https://t.co/zQkYmQHXfR

[SANS DFIR @sansforensics]

Toppling the Stack: Outlier Detection by @davidjbianco #threathunting #threathuntingsummit 2017 TALK NOW... https://t.co/iW29mgS9DN

[Dimiter Milushev @haskool]

Merry Christmas everyone!

[/Fay-lee-nuh/ @Felienne]

My plane was delayed, so this is the setup for my #fsharpconf talk in 20 minutes... #hoorayforairportwifi