Author name cluster

Chenhao Tan

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

10 papers

2 author rows

NeurIPS Conference 2025 Conference Paper

Absence Bench: Language Models Can’t See What’s Missing

Harvey Yiyun Fu
Aryan Shrivastava
Jared Moore
Peter West
Chenhao Tan
Ari Holtzman

Large language models (LLMs) are increasingly capable of processing long inputs and locating specific information within them, as evidenced by their performance on the Needle in a Haystack (NIAH) test. However, while models excel at recalling surprising information, they still struggle to identify clearly omitted information. We introduce AbsenceBench to assesses LLMs' capacity to detect missing information across three domains: numerical sequences, poetry, and GitHub pull requests. AbsenceBench asks models to identify which pieces of a document were deliberately removed, given access to both the original and edited contexts. Despite the apparent straightforwardness of these tasks, our experiments reveal that even state-of-the-art models like Claude-3. 7-Sonnet achieve only 69. 6% F1-score with a modest average context length of 5K tokens. Our analysis suggests this poor performance stems from a fundamental limitation: Transformer attention mechanisms cannot easily attend to "gaps" in documents since these absences don't correspond to any specific keys that can be attended to. Overall, our results and analysis provide a case study of the close proximity of tasks where models are already superhuman (NIAH) and tasks where models breakdown unexpectedly (AbsenceBench).

PDF Details

NeurIPS Conference 2025 Conference Paper

Concept Incongruence: An Exploration of Time and Death in Role Playing

Xiaoyan Bai
Ike Peng
Aditya Singh
Chenhao Tan

Consider this prompt "Draw a unicorn with two horns". Should large language models (LLMs) recognize that a unicorn has only one horn by definition and ask users for clarifications, or proceed to generate something anyway? We introduce concept incongruence to capture such phenomena where concept boundaries clash with each other, either in user prompts or in model representations, often leading to under-specified or mis-specified behaviors. In this work, we take the first step towards defining and analyzing model behavior under concept incongruence. Focusing on temporal boundaries in the Role-Play setting, we propose three behavioral metrics---abstention rate, conditional accuracy, and answer rate---to quantify model behavior under incongruence due to the role's death. We show that models fail to abstain after death and suffer from an accuracy drop compared to the Non-Role-Play setting. Through probing experiments, we identify two main causes: (i) unreliable encoding of the "death" state across different years, leading to unsatisfactory abstention behavior, and (ii) role playing causes shifts in the model’s temporal representations, resulting in accuracy drops. We leverage these insights to improve consistency in the model's abstention and answer behaviors. Our findings suggest that concept incongruence leads to unexpected model behaviors and point to future directions on improving model behavior under concept incongruence.

PDF Details

NeurIPS Conference 2025 Conference Paper

Prompting as Scientific Inquiry

Ari Holtzman
Chenhao Tan

Prompting is the primary method by which we study and control large language models. It is also one of the most powerful: nearly every major capability attributed to LLMs—few-shot learning, chain-of-thought, constitutional AI—was first unlocked through prompting. Yet prompting is rarely treated as science and is frequently frowned upon as alchemy. We argue that this is a category error. If we treat LLMs as a new kind of organism—complex, opaque, and trained rather than programmed—then prompting is not a workaround. It is behavioral science. Mechanistic interpretability peers into the neural substrate, prompting probes the model in its native interface: language. We argue that prompting is not inferior, but rather a key component in the science of LLMs.

PDF Details

TMLR Journal 2024 Journal Article

Causal Reasoning and Large Language Models: Opening a New Frontier for Causality

Emre Kiciman
Robert Ness
Amit Sharma
Chenhao Tan

The causal capabilities of large language models (LLMs) are a matter of significant debate, with critical implications for the use of LLMs in societally impactful domains such as medicine, science, law, and policy. We conduct a "behavorial" study of LLMs to benchmark their capability in generating causal arguments. Across a wide range of tasks, we find that LLMs can generate text corresponding to correct causal arguments with high probability, surpassing the best-performing existing methods. Algorithms based on GPT-3.5 and 4 outperform existing algorithms on a pairwise causal discovery task (97%, 13 points gain), counterfactual reasoning task (92%, 20 points gain) and event causality (86% accuracy in determining necessary and sufficient causes in vignettes). We perform robustness checks across tasks and show that the capabilities cannot be explained by dataset memorization alone, especially since LLMs generalize to novel datasets that were created after the training cutoff date. That said, LLMs exhibit unpredictable failure modes and we discuss the kinds of errors that may be improved and what are the fundamental limits of LLM-based answers. Overall, by operating on the text metadata, LLMs bring capabilities so far understood to be restricted to humans, such as using collected knowledge to generate causal graphs or identifying background causal context from natural language. As a result, LLMs may be used by human domain experts to save effort in setting up a causal analysis, one of the biggest impediments to the widespread adoption of causal methods. Given that LLMs ignore the actual data, our results also point to a fruitful research direction of developing algorithms that combine LLMs with existing causal techniques. Code and datasets are available at https://github.com/py-why/pywhy-llm.

PDF Details

NeurIPS Conference 2023 Conference Paper

Language Models Can Improve Event Prediction by Few-Shot Abductive Reasoning

Xiaoming Shi
Siqiao Xue
Kangrui Wang
Fan Zhou
James Zhang
Jun Zhou
Chenhao Tan
Hongyuan Mei

Large language models have shown astonishing performance on a wide range of reasoning tasks. In this paper, we investigate whether they could reason about real-world events and help improve the prediction performance of event sequence models. We design LAMP, a framework that integrates a large language model in event prediction. Particularly, the language model performs abductive reasoning to assist an event sequence model: the event model proposes predictions on future events given the past; instructed by a few expert-annotated demonstrations, the language model learns to suggest possible causes for each proposal; a search module finds out the previous events that match the causes; a scoring function learns to examine whether the retrieved events could actually cause the proposal. Through extensive experiments on several challenging real-world datasets, we demonstrate that our framework---thanks to the reasoning capabilities of large language models---could significantly outperform the state-of-the-art event sequence models.

PDF Details

ICLR Conference 2023 Conference Paper

Learning Human-Compatible Representations for Case-Based Decision Support

Han Liu
Yizhou Tian
Chacha Chen
Shi Feng 0005
Yuxin Chen 0001
Chenhao Tan

Algorithmic case-based decision support provides examples to help human make sense of predicted labels and aid human in decision-making tasks. Despite the promising performance of supervised learning, representations learned by supervised models may not align well with human intuitions: what models consider as similar examples can be perceived as distinct by humans. As a result, they have limited effectiveness in case-based decision support. In this work, we incorporate ideas from metric learning with supervised learning to examine the importance of alignment for effective decision support. In addition to instance-level labels, we use human-provided triplet judgments to learn human-compatible decision-focused representations. Using both synthetic data and human subject experiments in multiple classification tasks, we demonstrate that such representation is better aligned with human perception than representation solely optimized for classification. Human-compatible representations identify nearest neighbors that are perceived as more similar by humans and allow humans to make more accurate predictions, leading to substantial improvements in human decision accuracies (17.8% in butterfly vs. moth classification and 13.2% in pneumonia classification).

Details

TMLR Journal 2023 Journal Article

Machine Explanations and Human Understanding

Chacha Chen
Shi Feng
Amit Sharma
Chenhao Tan

Explanations are hypothesized to improve human understanding of machine learning models and achieve a variety of desirable outcomes, ranging from model debugging to enhancing human decision making. However, empirical studies have found mixed and even negative results. An open question, therefore, is under what conditions explanations can improve human understanding and in what way. To address this question, we first identify three core concepts that cover most existing quantitative measures of understanding: task decision boundary, model decision boundary, and model error. Using adapted causal diagrams, we provide a formal characterization of the relationship between these concepts and human approximations (i.e., understanding) of them. The relationship varies by the level of human intuition in different task types, such as emulation and discovery, which are often ignored when building or evaluating explanation methods. Our key result is that human intuitions are necessary for generating and evaluating machine explanations in human-AI decision making: without assumptions about human intuitions, explanations may improve human understanding of model decision boundary, but cannot improve human understanding of task decision boundary or model error. To validate our theoretical claims, we conduct human subject studies to show the importance of human intuitions. Together with our theoretical contributions, we provide a new paradigm for designing behavioral studies towards a rigorous view of the role of machine explanations across different tasks of human-AI decision making.

PDF Details

NeurIPS Conference 2022 Conference Paper

Probing Classifiers are Unreliable for Concept Removal and Detection

Abhinav Kumar
Chenhao Tan
Amit Sharma

Neural network models trained on text data have been found to encode undesirable linguistic or sensitive concepts in their representation. Removing such concepts is non-trivial because of a complex relationship between the concept, text input, and the learnt representation. Recent work has proposed post-hoc and adversarial methods to remove such unwanted concepts from a model's representation. Through an extensive theoretical and empirical analysis, we show that these methods can be counter-productive: they are unable to remove the concepts entirely, and in the worst case may end up destroying all task-relevant features. The reason is the methods' reliance on a probing classifier as a proxy for the concept. Even under the most favorable conditions for learning a probing classifier when a concept's relevant features in representation space alone can provide 100% accuracy, we prove that a probing classifier is likely to use non-concept features and thus post-hoc or adversarial methods will fail to remove the concept correctly. These theoretical implications are confirmed by experiments on models trained on synthetic, Multi-NLI, and Twitter datasets. For sensitive applications of concept removal such as fairness, we recommend caution against using these methods and propose a spuriousness metric to gauge the quality of the final classifier.

PDF Details

NeurIPS Conference 2019 Conference Paper

Ask not what AI can do, but what AI should do: Towards a framework of task delegability

Brian Lubars
Chenhao Tan

While artificial intelligence (AI) holds promise for addressing societal challenges, issues of exactly which tasks to automate and to what extent to do so remain understudied. We approach this problem of task delegability from a human-centered perspective by developing a framework on human perception of task delegation to AI. We consider four high-level factors that can contribute to a delegation decision: motivation, difficulty, risk, and trust. To obtain an empirical understanding of human preferences in different tasks, we build a dataset of 100 tasks from academic papers, popular media portrayal of AI, and everyday life, and administer a survey based on our proposed framework. We find little preference for full AI control and a strong preference for machine-in-the-loop designs, in which humans play the leading role. Among the four factors, trust is the most correlated with human preferences of optimal human-machine delegation. This framework represents a first step towards characterizing human preferences of AI automation across tasks. We hope this work encourages future efforts towards understanding such individual attitudes; our goal is to inform the public and the AI research community rather than dictating any direction in technology development.

PDF Details

AAAI Conference 2018 Conference Paper

Urban Dreams of Migrants: A Case Study of Migrant Integration in Shanghai

Yang Yang
Chenhao Tan
Zongtao Liu
Fei Wu
Yueting Zhuang

Unprecedented human mobility has driven the rapid urbanization around the world. In China, the fraction of population dwelling in cities increased from 17. 9% to 52. 6% between 1978 and 2012. Such large-scale migration poses challenges for policymakers and important questions for researchers. To investigate the process of migrant integration, we employ a one-month complete dataset of telecommunication metadata in Shanghai with 54 million users and 698 million call logs. We ﬁnd systematic differences between locals and migrants in their mobile communication networks and geographical locations. For instance, migrants have more diverse contacts and move around the city with a larger radius than locals after they settle down. By distinguishing new migrants (who recently moved to Shanghai) from settled migrants (who have been in Shanghai for a while), we demonstrate the integration process of new migrants in their ﬁrst three weeks. Moreover, we formulate classiﬁcation problems to predict whether a person is a migrant. Our classiﬁer is able to achieve an F1-score of 0. 82 when distinguishing settled migrants from locals, but it remains challenging to identify new migrants because of class imbalance. This classiﬁcation setup holds promise for identifying new migrants who will successfully integrate into locals (new migrants that misclassiﬁed as locals).

PDF Details