29 citations · 41 across the 22 of their papers we have counts for
10 papers · 1 filter
Fantastic Bugs and Where to Find Them in AI Benchmarks
Sang Truong, Yuheng Tu, Michael Hardy +8
Benchmarks are pivotal in driving AI progress, and invalid benchmark questions frequently undermine their reliability. Manually identifying and correcting errors among thousands of…
Towards Self-Refinement of Vision-Language Models with Triangular Consistency
Yunlong Deng, Guangyi Chen, Tianpei Gu +4
Vision-Language Models (VLMs) integrate visual knowledge with the analytical capabilities of Large Language Models (LLMs) through supervised visual instruction tuning, using image-…
Selection, Reflection and Self-Refinement: Revisit Reasoning Tasks via a Causal Lens
Yunlong Deng, Boyang Sun, Yan Li +4
Due to their inherent complexity, reasoning tasks have long been regarded as rigorous benchmarks for assessing the capabilities of machine learning models, especially large languag…
Position: Beyond Sensitive Attributes, ML Fairness Should Quantify Structural Injustice via Social Determinants
Zeyu Tang, Alex John London, Atoosa Kasirzadeh +4
Algorithmic fairness research has largely framed unfairness as discrimination along sensitive attributes. However, this approach limits visibility into unfairness as structural inj…
SmartCLIP: Modular Vision-language Alignment with Identification Guarantees
Shaoan Xie, Lingjing Kong, Yujia Zheng +5
Contrastive Language-Image Pre-training (CLIP)~\citep{radford2021learning} has emerged as a pivotal model in computer vision and multimodal learning, achieving state-of-the-art per…
Should Bias be Eliminated? A General Framework to Use Bias for OOD Generalization
Yan Li, Yunlong Deng, Zijian Li +4
Most approaches to out-of-distribution (OOD) generalization learn domain-invariant representations by discarding contextual bias. In this paper, we raise a critical question: Shoul…
Position: Mechanistic Interpretability Should Prioritize Feature Consistency in SAEs
Xiangchen Song, Aashiq Muhamed, Yujia Zheng +5
Sparse Autoencoders (SAEs) are a prominent tool in mechanistic interpretability (MI) for decomposing neural network activations into interpretable features. However, the aspiration…
When Selection Meets Intervention: Additional Complexities in Causal Discovery
Haoyue Dai, Ignavier Ng, Jianle Sun +5
We address the common yet often-overlooked selection bias in interventional studies, where subjects are selectively enrolled into experiments. For instance, participants in a drug…
Reflection-Window Decoding: Text Generation with Selective Refinement
Zeyu Tang, Zhenhao Chen, Xiangchen Song +6
The autoregressive decoding for text generation in large language models (LLMs), while widely used, is inherently suboptimal due to the lack of a built-in mechanism to perform refi…
Controllable Video Generation with Provable Disentanglement
Yifan Shen, Peiyuan Zhu, Zijian Li +6
Controllable video generation remains a significant challenge, despite recent advances in generating high-quality and consistent videos. Most existing methods for controlling video…