no code implementations • 27 Feb 2024 • Zhenting Qi, HANLIN ZHANG, Eric Xing, Sham Kakade, Himabindu Lakkaraju
Retrieval-Augmented Generation (RAG) improves pre-trained models by incorporating external knowledge at test time to enable customized adaptation.
no code implementations • 7 Dec 2023 • HANLIN ZHANG, Yi-Fan Zhang, Yaodong Yu, Dhruv Madeka, Dean Foster, Eric Xing, Himabindu Lakkaraju, Sham Kakade
Accurate uncertainty quantification is crucial for the safe deployment of machine learning models, and prior research has demonstrated improvements in the calibration of modern language models (LMs).
1 code implementation • 7 Nov 2023 • HANLIN ZHANG, Benjamin L. Edelman, Danilo Francati, Daniele Venturi, Giuseppe Ateniese, Boaz Barak
To prove this result, we introduce a generic efficient watermark attack; the attacker is not required to know the private key of the scheme or even which scheme is used.
no code implementations • 18 Sep 2023 • HANLIN ZHANG, Wenzheng Cheng
Compared to other methods for predicting the essentiality of lncRNA genes, our DeepHEN model not only tells whether sequence features or network spatial features have a greater influence on essentiality but also addresses the overfitting issue of those methods caused by the low number of essential lncRNA genes, as evidenced by the results of enrichment analysis.
1 code implementation • 5 May 2023 • HANLIN ZHANG, Jiani Huang, Ziyang Li, Mayur Naik, Eric Xing
We propose DSR-LM, a Differentiable Symbolic Reasoning framework where pre-trained LMs govern the perception of factual knowledge, and a symbolic module performs deductive reasoning.
1 code implementation • 6 Apr 2023 • Alexander Pan, Jun Shern Chan, Andy Zou, Nathaniel Li, Steven Basart, Thomas Woodside, Jonathan Ng, HANLIN ZHANG, Scott Emmons, Dan Hendrycks
And how do we measure these behaviors in general-purpose models such as GPT-4?
2 code implementations • 27 Mar 2023 • Xiangyuan Yang, Jie Lin, HANLIN ZHANG, Xinyu Yang, Peng Zhao
Although considerable efforts have been developed on improving the transferability of adversarial examples generated by transfer-based adversarial attacks, our investigation found that, the big deviation between the actual and steepest update directions of the current transfer-based adversarial attacks is caused by the large update step length, resulting in the generated adversarial examples can not converge well.
no code implementations • 17 Mar 2023 • Xiangyuan Yang, Jie Lin, HANLIN ZHANG, Xinyu Yang, Peng Zhao
In this paper, we first systematically investigated this issue and found that the enormous difference of attack success rates between the surrogate model and victim model is caused by the existence of a special area (known as fuzzy domain in our paper), in which the adversarial examples in the area are classified wrongly by the surrogate model while correctly by the victim model.
1 code implementation • 16 Dec 2022 • Yi-Fan Zhang, HANLIN ZHANG, Li Erran Li, Eric Xing
Pre-trained language models (LMs) have shown remarkable reasoning performance using explanations or chain-of-thoughts (CoT)) for in-context learning.
no code implementations • 15 Oct 2022 • HANLIN ZHANG, Xuechen Li, Prithviraj Sen, Salim Roukos, Tatsunori Hashimoto
Across 7 tasks, temperature scaling and Platt scaling with DP-SGD result in an average 3. 1-fold reduction in the in-domain expected calibration error and only incur at most a minor percent drop in accuracy.
no code implementations • 2 Jun 2022 • Xiangyuan Yang, Jie Lin, HANLIN ZHANG, Xinyu Yang, Peng Zhao
To enhance the robustness of the classifier, in our paper, a \textbf{F}eature \textbf{A}nalysis and \textbf{C}onditional \textbf{M}atching prediction distribution (FACM) model is proposed to utilize the features of intermediate layers to correct the classification.
no code implementations • 2 Jun 2022 • Xiangyuan Yang, Jie Lin, HANLIN ZHANG, Xinyu Yang, Peng Zhao
The empirical and theoretical analysis demonstrates that the MDL loss improves the robustness and generalization of the model simultaneously for natural training.
no code implementations • 19 May 2022 • Xiangyuan Yang, Jie Lin, HANLIN ZHANG, Xinyu Yang, Peng Zhao
Specifically, we propose a gradient aligned mechanism to ensure that the derivatives of the loss function with respect to the logit vector have the same weight coefficients between the surrogate and victim models.
1 code implementation • 2 Feb 2022 • Yi-Fan Zhang, HANLIN ZHANG, Zachary C. Lipton, Li Erran Li, Eric P. Xing
Previous works on Treatment Effect Estimation (TEE) are not in widespread use because they are predominantly theoretical, where strong parametric assumptions are made but untractable for practical application.
no code implementations • 30 Jan 2022 • Liu Ziyin, HANLIN ZHANG, Xiangming Meng, Yuting Lu, Eric Xing, Masahito Ueda
This work theoretically studies stochastic neural networks, a main type of neural network in use.
1 code implementation • CVPR 2022 • HANLIN ZHANG, Yi-Fan Zhang, Weiyang Liu, Adrian Weller, Bernhard Schölkopf, Eric P. Xing
To tackle this challenge, we first formalize the OOD generalization problem as constrained optimization, called Disentanglement-constrained Domain Generalization (DDG).
1 code implementation • 5 Nov 2021 • Haohan Wang, Zeyi Huang, HANLIN ZHANG, Yong Jae Lee, Eric Xing
Machine learning has demonstrated remarkable prediction accuracy over i. i. d data, but the accuracy often drops when tested with data from another distribution.
1 code implementation • NeurIPS 2020 • Wangchunshu Zhou, Jinyi Hu, HANLIN ZHANG, Xiaodan Liang, Maosong Sun, Chenyan Xiong, Jian Tang
In this paper, we develop a general framework for interpretable natural language understanding that requires only a small set of human annotated explanations for training.
no code implementations • 23 Oct 2020 • HANLIN ZHANG, Shuai Lin, Weiyang Liu, Pan Zhou, Jian Tang, Xiaodan Liang, Eric P. Xing
Recently, there has been increasing interest in the challenge of how to discriminatively vectorize graphs.
no code implementations • journal 2018 • XINRUI GE, JIA YU, Chengyu Hu, HANLIN ZHANG, AND RONG HAO
In searchable encryption, the cloud server might return the invalid result to data user for saving the computation cost or other reasons.