Information Elicitation Meets Large Language Models
Published:
Posts by Collection
portfolio
projects
publications
Equal Affection or Random Selection: the Quality of Subjective Feedback from a Group Perspective
Published in Arxiv, 2021
This paper designs a family of group level evaluation metrics expressing informativeness, based on an additionally asking for predictions about others.
Download here
SURPRISE! and When to Schedule It
Published in the 30th International Joint Conference on Artificial Intelligence (IJCAI), 2021
This experimental paper demonstrates that audience utility has some properties by organizing respondents to watch LOL S10 and report information.
Download here
FileInsurer: A Scalable and Reliable Protocol for Decentralized File Storage in Blockchain
Published in the 42nd IEEE International Conference on Distributed Computing Systems (ICDCS), 2022
This paper proposes a secure and scalable design of decentralized storage network, named FileInsurer.
Download here
Empirical Analysis of EIP-1559: Transaction Fees, Waiting Time, and Consensus Security
Published in the 29th ACM Conference on Computer and Communications Security (CCS), 2022
This paper is about the empirical analysis of EIP-1559.
Download here
A Framework of Transaction Packaging in High-throughput Blockchains
Published in Arxiv, 2023
This paper propose a framework of transaction packaging problem in High-throughput Blockchains.
Download here
Calibrating “Cheap Signals” in Peer Review without a Prior
Published in the 37th Conference on Neural Information Processing Systems (NeurIPS), 2023
This paper propose a novel method to calibrate systematic bias in peer review.
Download here
How Gold to Make the Golden Snitch: Designing the “Game Changer” in Esports
Published in ArXiv, 2024
Game-theoretic and empirical analysis of reward design for a game-changing esports item.
Download here
Eliciting Informative Text Evaluations with Large Language Models
Published in 2024 ACM Conference on Economics and Computation (EC), 2024
This paper propose a method to elicit informative text evaluations leveraging the power of LLMs.
Download here
Benchmarking LLMs’ Judgments with No Gold Standard
Published in the 13th International Conference on Learning Representations (ICLR), 2025
A mutual-information-based metric for benchmarking LLM judgments without ground-truth labels.
Download here
Aligned Textual Scoring Rules
Published in ArXiv, 2025
This paper propose a novel method to create human-aligned proper scoring rules.
Download here
Making and Evaluating Calibrated Forecasts
Published in ArXiv, 2025
Designs truthful calibration metrics for evaluating model forecasts.
Download here
Jailbreaking LLMs via Calibration
Published in ArXiv, 2026
Reframes jailbreak attacks as calibration and proposes a gradient-shift attack.
Download here
talks
teaching
Algorithm Design and Analysis
TA, Peking University, Department of computer science and technology, 2021
Mathematical Foundations for the Information Age
TA, Peking University, Department of computer science and technology, 2021
Mathematical Foundations for the Information Age
TA, Peking University, Department of computer science and technology, 2022