论文检索

Miscellaneous Aspects of Machine Learning/Causality

Hao Wang, Jiajun Fan, Zhichao Chen, Haoxuan Li, Weiming Liu, Tianqiao Liu, Quanyu Dai, Yichao Wang, Zhenhua Dong, Ruiming Tang

Estimating individual treatment effects from observational data is challenging due to treatment selection bias. Prevalent methods mainly mitigate this issue by aligning different treatment groups in the latent space, the core of which is the calculation of distribution discrepancy. However, two issues that are often overlooked can render these methods invalid:(1) mini-batch sampling effects (MSE), where the calculated discrepancy is erroneous in non-ideal mini-batches with outcome imbalance and outliers;(2) unobserved confounder effects (UCE), where the unobserved confounders are not considered in the discrepancy calculation.Both of these issues invalidate the calculated discrepancy, mislead the training of estimators, and thus impede the handling of treatment selection bias.To tackle these issues, we propose Entire Space CounterFactual Regression (ESCFR), which is a new take on optimal transport technology in the context of causality.Specifically, based on the canonical optimal transport framework, we propose a relaxed mass-preserving regularizer to address the MSE issue and design a proximal factual outcome regularizer to handle the UCE issue.Extensive experiments demonstrate that ESCFR estimates distribution discrepancy accurately, handles the treatment selection bias effectively, and outperforms prevalent competitors significantly.

Applications/Language, Speech and Dialog

Zeqiu Wu, Yushi Hu, Weijia Shi, Nouha Dziri, Alane Suhr, Prithviraj Ammanabrolu, Noah Smith, Mari Ostendorf, Hannaneh Hajishirzi

Language models (LMs) often exhibit undesirable text generation behaviors, including generating false, toxic, or irrelevant outputs. Reinforcement learning from human feedback (RLHF)---where human preference judgments on LM outputs are transformed into a learning signal---has recently shown promise in addressing these issues. However, such holistic feedback conveys limited information on long text outputs; it does not indicate which aspects of the outputs influenced user preference; e.g., which parts contain what type(s) of errors. In this paper, we use fine-grained human feedback (e.g., which sentence is false, which sub-sentence is irrelevant) as an explicit training signal. We introduce Fine-Grained RLHF, a framework that enables training and learning from reward functions that are fine-grained in two respects: (1) density, providing a reward after every segment (e.g., a sentence) is generated; and (2) incorporating multiple reward models associated with different feedback types (e.g., factual incorrectness, irrelevance, and information incompleteness). We conduct experiments on detoxification and long-form question answering to illustrate how learning with this reward function leads to improved performance, supported by both automatic and human evaluation. Additionally, we show that LM behaviors can be customized using different combinations of fine-grained reward models. We release all data, collected human feedback, and codes at https://FineGrainedRLHF.github.io.

Applications/Language, Speech and Dialog

Yuheng Zha, Yichi Yang, Ruichen Li, Zhiting Hu

Large language models (LLMs), typically designed as a function of next-word prediction, have excelled across extensive NLP tasks. Despite the generality, next-word prediction is often not an efficient formulation for many of the tasks, demanding an extreme scale of model parameters (10s or 100s of billions) and sometimes yielding suboptimal performance.In practice, it is often desirable to build more efficient models---despite being less versatile, they still apply to a substantial subset of problems, delivering on par or even superior performance with much smaller model sizes.In this paper, we propose text alignment as an efficient unified model for a wide range of crucial tasks involving text entailment, similarity, question answering (and answerability), factual consistency, and so forth. Given a pair of texts, the model measures the degree of alignment between their information. We instantiate an alignment model through lightweight finetuning of RoBERTa (355M parameters) using 5.9M examples from 28 datasets. Despite its compact size, extensive experiments show the model's efficiency and strong performance: (1) On over 20 datasets of aforementioned diverse tasks, the model matches or surpasses FLAN-T5 models that have around 2x or 10x more parameters; the single unified model also outperforms task-specific models finetuned on individual datasets; (2) When applied to evaluate factual consistency of language generation on 23 datasets, our model improves over various baselines, including the much larger GPT-3.5 (ChatGPT) and sometimes even GPT-4; (3) The lightweight model can also serve as an add-on component for LLMs such as GPT-3.5 in question answering tasks, improving the average exact match (EM) score by 17.94 and F1 score by 15.05 through identifying unanswerable questions.

Social Aspects/Accountability, Transparency and Interpretability

Peter Hase, Mohit Bansal, Been Kim, Asma Ghandeharioun

Language models learn a great quantity of factual information during pretraining, and recent work localizes this information to specific model weights like mid-layer MLP weights. In this paper, we find that we can change how a fact is stored in a model by editing weights that are in a different location than where existing methods suggest that the fact is stored. This is surprising because we would expect that localizing facts to specific model parameters would tell us where to manipulate knowledge in models, and this assumption has motivated past work on model editing methods. Specifically, we show that localization conclusions from representation denoising (also known as Causal Tracing) do not provide any insight into which model MLP layer would be best to edit in order to override an existing stored fact with a new one. This finding raises questions about how past work relies on Causal Tracing to select which model layers to edit. Next, we consider several variants of the editing problem, including erasing and amplifying facts. For one of our editing problems, editing performance does relate to localization results from representation denoising, but we find that which layer we edit is a far better predictor of performance. Our results suggest, counterintuitively, that better mechanistic understanding of how pretrained language models work may not always translate to insights about how to best change their behavior.

Guangyi Chen, Junlong Li, Jiwen Lu, Jie Zhou

Forecasting human trajectories in complex dynamic environments plays a critical role in autonomous vehicles and intelligent robots. Most existing methods learn to predict future trajectories by behavior clues from history trajectories and interaction clues from environments. However, the inherent bias between training and deployment environments is ignored. Hence, we propose a counterfactual analysis method for human trajectory prediction to investigate the causality between the predicted trajectories and input clues and alleviate the negative effects brought by environment bias. We first build a causal graph for trajectory forecasting with history trajectory, future trajectory, and the environment interactions. Then, we cut off the inference from the environment to trajectory by constructing the counterfactual intervention on the trajectory itself. Finally, we compare the factual and counterfactual trajectory clues to alleviate the effects of environment bias and highlight the trajectory clues. Our counterfactual analysis is a plug-and-play module that can be applied to any baseline prediction methods including RNN- and CNN-based ones. We show that our method achieves consistent improvement for different baselines and obtains state-of-the-art results on public pedestrian trajectory forecasting benchmarks.

Hareesh Ravi, Kushal Kafle, Scott Cohen, Jonathan Brandt, Mubbasir Kapadia

Visual storytelling and story comprehension are uniquely human skills that play a central role in how we learn about and experience the world. Despite remarkable progress in recent years in synthesis of visual and textual content in isolation and learning effective joint visual-linguistic representations, existing systems still operate only at a superficial, factual level. With the goal of developing systems that are able to comprehend rich human-generated narratives, and co-create new stories, we introduce AESOP: a new dataset that captures the creative process associated with visual storytelling. Visual panels are composed of clip-art objects with specific attributes enabling a broad range of creative expression. Using AESOP, we propose foundational storytelling tasks that are generative variants of story cloze tests, to better measure the creative and causal reasoning ability required for visual storytelling. We further develop a generalized story completion framework that models stories as the co-evolution of visual and textual concepts. We benchmark the proposed approach with human baselines and evaluate using comprehensive qualitative and quantitative metrics. Our results highlight key insights related to the dataset, modelling and evaluation of visual storytelling for future research in this promising field of study.