Haosen Ge, Ph.D.

22 posts

Haosen Ge, Ph.D. banner
Haosen Ge, Ph.D.

Haosen Ge, Ph.D.

@HaosenGe

Data Scientist @WhartonAIAI | Ph.D. from @Princeton

Katılım Aralık 2017
97 Takip Edilen100 Takipçiler
Haosen Ge, Ph.D. retweetledi
Hamsa Bastani
Hamsa Bastani@hamsabastani·
Conformal prediction sets are a useful way to capture uncertainty for LLMs & deep learning models. But they're data-hungry! We propose a semi-bandit algo to learn these sets online. Check out our @icmlconf paper: arxiv.org/abs/2405.13268 Work led by @HaosenGe & with @obastani
English
1
4
13
1.2K
Kan Xu
Kan Xu@kanxu526·
🚨 New Paper Alert🚨 I’m very happy to announce that my job market paper is now official on【Management Science】! We propose a novel multitask learning & bandit algorithm built on robust statistics and Lasso, which shows notable performance both theoretically and practically.
Kan Xu tweet media
English
3
3
26
1.7K
Haosen Ge, Ph.D. retweetledi
Handi Li
Handi Li@Handi_Rita_Li·
Excited to see Philip K.H. Wong Centre @HKUniversity for promoting my working paper! I examined how Open Government Information mobilizes legal resistance in China but suffers a potenial backlash from local governments. Welcome any comments!
Philip K.H. Wong Centre for Chinese Law@CCLHKU

Read this article by @Handi_Rita_Li highlighting the mixed effects of open government information in China. Chinese Law Blog: ccl.law.hku.hk/chinese-law-bl… Chinese Law eJournal: papers.ssrn.com/sol3/papers.cf…

English
0
3
24
2.5K
Haosen Ge, Ph.D.
Haosen Ge, Ph.D.@HaosenGe·
3/ To address this issue, we propose a class of new algorithms: compliance robust algorithms. We demonstrate that it is possible to convert any machine learning model into compliance robust models that guarantee improvement in outcome accuracy and fairness.
English
0
0
0
55
Haosen Ge, Ph.D.
Haosen Ge, Ph.D.@HaosenGe·
It's fantastic to see the increased interest in our research! I firmly believe it addresses a crucial question that any organization looking to implement AI to support human decision-making should consider.
Wharton AI & Analytics Initiative@WhartonAIAI

.@Wharton's@hamsabastani, Haosen Ge from @WhartonAnalytics, & @obastani explore AI-human collaboration challenges in their latest research featured in @Medium's "How to Improve AI Fairness in an Unfair World." Check it out here: whr.tn/3VRkuys

English
2
0
9
575
Haosen Ge, Ph.D.
Haosen Ge, Ph.D.@HaosenGe·
2/In a nutshell, our research highlights a key issue with current AI decision support systems: the potential for AI-assisted humans to perform worse than either AI operating alone or humans without AI. This contradicts the very purpose of integrating AI systems.
English
0
0
0
57
Haosen Ge, Ph.D.
Haosen Ge, Ph.D.@HaosenGe·
Excited to present my work with @hamsabastani and @obastani at the Simons Institute for the Theory of Computing (@SimonsInstitute) ! We delve into the challenges with algorithmic fairness when dealing with human decision makers who can selectively comply.
Haosen Ge, Ph.D. tweet mediaHaosen Ge, Ph.D. tweet media
English
1
2
19
2K
Haosen Ge, Ph.D. retweetledi
Hamsa Bastani
Hamsa Bastani@hamsabastani·
If you play chess, consider signing up for the study 😊🙏
International Chess Federation@FIDE_chess

INSEAD (@INSEAD) and Wharton (@Wharton) are conducting a study on AI and algorithmic advice in chess. The study aims to provide insights into how AI and algorithmic advice and their effect on sequential human decision-making. The research, grounded in chess, will offer valuable knowledge to businesses, academia, and the chess community. 🔎 The experiment is fun, lasts about 2-3 hours, and you get to play chess against their custom-made, Stockfish-based chess engine that adjusts its personality to a player's rating/skill. ♟️ Each participant gets compensated according to their rating, and the only requirement for someone to participate is to have either a FIDE rating from 1000-2699. Registration is open now! Sign up using the link below 👇 wharton.qualtrics.com/jfe/form/SV_e5…

English
0
3
11
2.1K
In Young Park
In Young Park@in_young_park·
So inspired by @ProfZhaoMIT's presentation today at @BU_MSE. One of the best talks I've been to in recent years. Looking forward to following the bioadhesive ultrasound space in the future!
English
1
0
9
3.1K
Rory Truex
Rory Truex@rorytruex·
Grateful for the many, many people who helped me get to this point-- especially those that provided support during some of the harder times. I will never be able to pay them all back, but I hope to use the next stage of my career to pay it forward.
Princeton Politics@PUPolitics

Congratulations to @rorytruex for his promotion to Associate Professor of Politics and International Affairs!

English
86
3
528
120K
Haosen Ge, Ph.D. retweetledi
Erik Brynjolfsson
Erik Brynjolfsson@erikbryn·
Nothing to see here. Just an AI system learning to strategically plan world conquest, negotiate with humans in English and then betray them via the game of Diplomacy. science.org/doi/10.1126/sc…
English
27
262
1.3K
0
Haosen Ge, Ph.D. retweetledi
James Zou
James Zou@james_y_zou·
Given a trained #ML model, can we estimate how well it works on new data w/o seeing any labels? Can we still do it even if both covariates and labels shift in data? Our #NeurIPS2022 paper shows how to est model perf on new data in this hard case! arxiv.org/pdf/2209.08436… 1/2
James Zou tweet mediaJames Zou tweet media
English
4
46
189
0
Haosen Ge, Ph.D. retweetledi
Misha Laskin
Misha Laskin@MishaLaskin·
In our new work - Algorithm Distillation - we show that transformers can improve themselves autonomously through trial and error without ever updating their weights. No prompting, no finetuning. A single transformer collects its own data and maximizes rewards on new tasks. 1/N
English
22
241
1.3K
0

Keşfet

@icmlconf @obastani @kanxu526 @HKUniversity @sid_devic @SimonsInstitute @hamsabastani @StarCraft