Ariah Klages-Mundt

@aklamun

PhD student , applied math. I work on complex systems, including / design and network cascades.

medium.com/@aklamun
Vrijeme pridruživanja: rujan 2012.
9 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @aklamun

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @aklamun

  1. Prikvačeni tweet
    12. lip 2019.

    Check out my work on modeling noncustodial "(In)Stability for the Blockchain: Deleveraging Spirals and Stablecoin Attacks". A foundational step in a scientific approach to stablecoin design. I will present at June 25.

    7 replies 44 proslijeđena tweeta 144 korisnika označavaju da im se sviđa
    Poništi
  2. 31. sij

    The 2-week based estimator above isn't necessarily any better actually. We also don't know how long until an infection becomes fatal or not. And the limited data on that probably isn't easily accessible. All we can say is that the rate is *probably* >2%

    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  3. 31. sij

    My understanding is we basically have no idea what the actual death and recovery rates are this early. Anyone else have a better take on this? (4/4)

    1 korisnik označava da mu se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  4. 31. sij

    Less biased estimator would be current deaths/reported cases two weeks ago (~400): 200/400 = 50%. But the margin of error is huge with that sparse of data and considering that there is a lot of noise to the 'true' average rates (3/4)

    1 reply
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  5. 31. sij

    ...to report 200/10k = 2% as death rate since it could be weeks before patients die/recover. By the same measure, recovery rate is also 2%, which clearly doesn't make sense. The estimator is a biased underestimate by an exponential factor (2/4)

    1 reply
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  6. 31. sij

    The coronavirus death rate is commonly reported as 2-3%. But is this reasonable? Reported stats are ~10k infected patients, ~200 deaths, and ~200 recoveries. Since we're in the first few weeks of the outbreak, (1) the data is pretty sparse, and (2) it's pretty deceiving... (1/4)

    1 reply 1 proslijeđeni tweet 5 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  7. 31. sij

    I am honored to receive the PhD fellowship from Bloomberg. Many thanks for supporting my work

    Congratulations to the 2019-2020 Ph.D. Fellows: ' , 's Center for Applied Math's , ’s Yi Su, & 's . Learn more about them & their research:
    1 proslijeđeni tweet 19 korisnika označava da im se sviđa
    Poništi
  8. 31. sij

    This value accrual can be from a traditional sense of cashflows, from use cases providing natural demand, or more intangible things like value from decentralization/censorship resistance

    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  9. 31. sij

    Reminder: cashflow/value accrual expectations, whether honest or through attack, usually determine rational (e.g., game theory) valuations, not arbitrary number-go-up economics. Haven't seen a convincing argument that any system is an exception

    1 reply 3 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  10. proslijedio/la je Tweet
    11. sij

    (1/7) Here is a new paper on hypergraph cuts by , the twitterless Jon Kleinberg, and me that I am very excited about: min s-t graph cut is one of the basic problems we learn about in an algo class. What if we want min s-t _hypergraph_ cut?

    1 reply 16 proslijeđenih tweetova 78 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet
    17. pro 2019.

    BDoS: Blockchain Denial of Service. Attacker incentivizes miners to stop with far less than 51%, bringing a Nakamoto-like chain to a halt if miners are profit-driven. w/ , Ji, Pang, Klages-Mundt,

    8 replies 30 proslijeđenih tweetova 93 korisnika označavaju da im se sviđa
    Poništi
  12. 9. pro 2019.

    E.g., why should a governance contract that directly extracts the collateral and/or mints Dai without corresponding collateral ever have been in the set of valid governance contracts in the first place? (Although that doesn't exclude indirect ways to do such things...)

    4 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  13. 9. pro 2019.

    Independent of a time delay, it's also pretty surprising that there aren't basic constraints that limit the scope of a valid governance contract. Seems that there should be some level of hardcoding of the Dai design that MKR holders must work within

    I'm a bit late to my own party due to Medium timed publish and an irregular sleep cycle, but here it is!
    2 proslijeđena tweeta 13 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  14. 14. stu 2019.

    Check out my new post on and attacks. "Vulnerabilities in : Oracle-Governance Attacks, Attack , and (De)Centralization"

    7 replies 11 proslijeđenih tweetova 34 korisnika označavaju da im se sviđa
    Poništi
  15. proslijedio/la je Tweet
    8. lis 2019.

    Nice talk by on non-custodial stablecoins and possible attack scenarios

    1 proslijeđeni tweet 4 korisnika označavaju da im se sviđa
    Poništi
  16. 8. lis 2019.

    Come hear my talk at Devcon5 about crashes, risks, and attacks. Tomorrow (Oct 9) 11:35am

    3 proslijeđena tweeta 14 korisnika označava da im se sviđa
    Poništi
  17. 1. lis 2019.

    My talk on at is now online A high level overview of stablecoins and models with visual explanations Main take-away: feedback effects in can lead to faster collateral drawdown and perverse incentives for attack

    2 proslijeđena tweeta 11 korisnika označava da im se sviđa
    Poništi
  18. proslijedio/la je Tweet
    26. ruj 2019.

    This is why they tell you not to read your own book reviews. Thanks Howard – could you suggest a book where I could learn more about “chaos theory”?

    81 reply 142 proslijeđena tweeta 1.579 korisnika označava da im se sviđa
    Poništi
  19. proslijedio/la je Tweet
    24. ruj 2019.

    The price drop in ETH had some very interesting side effects for DAI and the CDP contracts that collateralize it.

    Due to the heavy drop in ETH price coinciding with network congestion and a spike in gas prices today, our automation system struggled to execute all needed CDP ratio adjustments in time. (1/4)
    Prikaži ovu nit
    21 reply 97 proslijeđenih tweetova 408 korisnika označava da im se sviđa
    Poništi
  20. 16. ruj 2019.

    A lot of risk in DeFi is ignored today. These are systems we don't understand very well. They're not necessarily safe just because they're on blockchain. We've already seen glimpses of crises in stablecoins. It's quite a challenge to design complex systems that work well

    Tweet je nedostupan.
    4 korisnika označavaju da im se sviđa
    Poništi
  21. proslijedio/la je Tweet
    14. ruj 2019.

    Interesting attack on EOS allows hacker to turn $1k into $110k.

    45 replies 276 proslijeđenih tweetova 708 korisnika označava da im se sviđa
    Poništi

@aklamun još nije objavio nijedan Tweet.

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·