Language Models as Causal Effect Generators

We present a framework for large language model (LLM) based data generation with controllable causal structure. In particular, we define a procedure for turning any language model and any directed acyclic graph (DAG) into a sequence-driven structural causal model (SD-SCM). Broadly speaking, an SD-SCM is a causal model with user-defined structure and LLM-defined structural equations. We characterize how an SD-SCM allows sampling from observational, interventional, and counterfactual distributions according to the desired causal structure. We then leverage this procedure to propose a new type of benchmark for causal inference methods, generating individual-level counterfactual data without needing to manually specify functional relationships between variables. We create an example benchmark consisting of thousands of datasets, and test a suite of popular estimation methods on these datasets for average, conditional average, and individual treatment effect estimation, both with and without hidden confounding. Apart from generating data, the same procedure also allows us to test for the presence of a causal effect that might be encoded in an LLM. This procedure can underpin auditing LLMs for misinformation, discrimination, or otherwise undesirable behavior. We believe SD-SCMs can serve as a useful tool in any application that would benefit from sequential data with controllable causal structure.

Aioli: A Unified Optimization Framework for Language Model Data Mixing

Language model performance depends on identifying the optimal mixture of data groups to train on (e.g., law, code, math). Prior work has proposed a diverse set of methods to efficiently learn mixture proportions, ranging from fitting regression models over training runs to dynamically updating proportions throughout training. Surprisingly, we find that no existing method consistently outperforms a simple stratified sampling baseline in terms of average test perplexity per group. In this paper, we study the cause of this inconsistency by unifying existing methods into a standard optimization framework. We show that all methods set proportions to minimize total loss, subject to a method-specific mixing law -- an assumption on how loss is a function of mixture proportions. We find that existing parameterizations of mixing laws can express the true loss-proportion relationship empirically, but the methods themselves often set the mixing law parameters inaccurately, resulting in poor and inconsistent performance. Finally, we leverage the insights from our framework to derive a new online method named Aioli, which directly estimates the mixing law parameters throughout training and uses them to dynamically adjust proportions. Empirically, Aioli outperforms stratified sampling on 6 out of 6 datasets by an average of 0.28 test perplexity points, whereas existing methods fail to consistently beat stratified sampling, doing up to 6.9 points worse. Moreover, in a practical setting where proportions are learned on shorter runs due to computational constraints, Aioli can dynamically adjust these proportions over the full training run, consistently improving performance over existing methods by up to 12.01 test perplexity points.

When Does Classical Chinese Help? Quantifying Cross-Lingual Transfer in Hanja and Kanbun

Historical and linguistic connections within the Sinosphere have led researchers to use Classical Chinese resources for cross-lingual transfer when processing historical documents from Korea and Japan. In this paper, we question the assumption of cross-lingual transferability from Classical Chinese to Hanja and Kanbun, the ancient written languages of Korea and Japan, respectively. Our experiments across machine translation, named entity recognition, and punctuation restoration tasks show minimal impact of Classical Chinese datasets on language model performance for ancient Korean documents written in Hanja, with performance differences within $\pm{}0.0068$ F1-score for sequence labeling tasks and up to $+0.84$ BLEU score for translation. These limitations persist consistently across various model sizes, architectures, and domain-specific datasets. Our analysis reveals that the benefits of Classical Chinese resources diminish rapidly as local language data increases for Hanja, while showing substantial improvements only in extremely low-resource scenarios for both Korean and Japanese historical documents. These mixed results emphasize the need for careful empirical validation rather than assuming benefits from indiscriminate cross-lingual transfer.

Semiparametric conformal prediction

Many risk-sensitive applications require well-calibrated prediction sets over multiple, potentially correlated target variables, for which the prediction algorithm may report correlated non-conformity scores. In this work, we treat the scores as random vectors and aim to construct the prediction set accounting for their joint correlation structure. Drawing from the rich literature on multivariate quantiles and semiparametric statistics, we propose an algorithm to estimate the $1-\alpha$ quantile of the scores, where $\alpha$ is the user-specified miscoverage rate. In particular, we flexibly estimate the joint cumulative distribution function (CDF) of the scores using nonparametric vine copulas and improve the asymptotic efficiency of the quantile estimate using its influence function. The vine decomposition allows our method to scale well to a large number of targets. We report desired coverage and competitive efficiency on a range of real-world regression problems, including those with missing-at-random labels in the calibration set.

LLMs are Highly-Constrained Biophysical Sequence Optimizers

Large language models (LLMs) have recently shown significant potential in various biological tasks such as protein engineering and molecule design. These tasks typically involve black-box discrete sequence optimization, where the challenge lies in generating sequences that are not only biologically feasible but also adhere to hard fine-grained constraints. However, LLMs often struggle with such constraints, especially in biological contexts where verifying candidate solutions is costly and time-consuming. In this study, we explore the possibility of employing LLMs as highly-constrained bilevel optimizers through a methodology we refer to as Language Model Optimization with Margin Expectation (LLOME). This approach combines both offline and online optimization, utilizing limited oracle evaluations to iteratively enhance the sequences generated by the LLM. We additionally propose a novel training objective -- Margin-Aligned Expectation (MargE) -- that trains the LLM to smoothly interpolate between the reward and reference distributions. Lastly, we introduce a synthetic test suite that bears strong geometric similarity to real biophysical problems and enables rapid evaluation of LLM optimizers without time-consuming lab validation. Our findings reveal that, in comparison to genetic algorithm baselines, LLMs achieve significantly lower regret solutions while requiring fewer test function evaluations. However, we also observe that LLMs exhibit moderate miscalibration, are susceptible to generator collapse, and have difficulty finding the optimal solution when no explicit ground truth rewards are available.

Generalizing to any diverse distribution: uniformity, gentle finetuning and rebalancing

As training datasets grow larger, we aspire to develop models that generalize well to any diverse test distribution, even if the latter deviates significantly from the training data. Various approaches like domain adaptation, domain generalization, and robust optimization attempt to address the out-of-distribution challenge by posing assumptions about the relation between training and test distribution. Differently, we adopt a more conservative perspective by accounting for the worst-case error across all sufficiently diverse test distributions within a known domain. Our first finding is that training on a uniform distribution over this domain is optimal. We also interrogate practical remedies when uniform samples are unavailable by considering methods for mitigating non-uniformity through finetuning and rebalancing. Our theory provides a mathematical grounding for previous observations on the role of entropy and rebalancing for o.o.d. generalization and foundation model training. We also provide new empirical evidence across tasks involving o.o.d. shifts which illustrate the broad applicability of our perspective.

Using Deep Autoregressive Models as Causal Inference Engines

Existing causal inference (CI) models are limited to primarily handling low-dimensional confounders and singleton actions. We propose an autoregressive (AR) CI framework capable of handling complex confounders and sequential actions common in modern applications. We accomplish this by {\em sequencification}, transforming data from an underlying causal diagram into a sequence of tokens. This approach not only enables training with data generated from any DAG but also extends existing CI capabilities to accommodate estimating several statistical quantities using a {\em single} model. We can directly predict interventional probabilities, simplifying inference and enhancing outcome prediction accuracy. We demonstrate that an AR model adapted for CI is efficient and effective in various complex applications such as navigating mazes, playing chess endgames, and evaluating the impact of certain keywords on paper acceptance rates.

Targeted Cause Discovery with Data-Driven Learning

We propose a novel machine learning approach for inferring causal variables of a target variable from observations. Our goal is to identify both direct and indirect causes within a system, thereby efficiently regulating the target variable when the difficulty and cost of intervening on each causal variable vary. Our method employs a neural network trained to identify causality through supervised learning on simulated data. By implementing a local-inference strategy, we achieve linear complexity with respect to the number of variables, efficiently scaling up to thousands of variables. Empirical results demonstrate the effectiveness of our method in identifying causal relationships within large-scale gene regulatory networks, outperforming existing causal discovery methods that primarily focus on direct causality. We validate our model's generalization capability across novel graph structures and generating mechanisms, including gene regulatory networks of E. coli and the human K562 cell line. Implementation codes are available at https://github.com/snu-mllab/Targeted-Cause-Discovery.

Analysis of the ICML 2023 Ranking Data: Can Authors' Opinions of Their Own Papers Assist Peer Review in Machine Learning?

We conducted an experiment during the review process of the 2023 International Conference on Machine Learning (ICML) that requested authors with multiple submissions to rank their own papers based on perceived quality. We received 1,342 rankings, each from a distinct author, pertaining to 2,592 submissions. In this paper, we present an empirical analysis of how author-provided rankings could be leveraged to improve peer review processes at machine learning conferences. We focus on the Isotonic Mechanism, which calibrates raw review scores using author-provided rankings. Our analysis demonstrates that the ranking-calibrated scores outperform raw scores in estimating the ground truth ``expected review scores'' in both squared and absolute error metrics. Moreover, we propose several cautious, low-risk approaches to using the Isotonic Mechanism and author-provided rankings in peer review processes, including assisting senior area chairs' oversight of area chairs' recommendations, supporting the selection of paper awards, and guiding the recruitment of emergency reviewers. We conclude the paper by addressing the study's limitations and proposing future research directions.

Non-convolutional Graph Neural Networks

Rethink convolution-based graph neural networks (GNN) -- they characteristically suffer from limited expressiveness, over-smoothing, and over-squashing, and require specialized sparse kernels for efficient computation. Here, we design a simple graph learning module entirely free of convolution operators, coined random walk with unifying memory (RUM) neural network, where an RNN merges the topological and semantic graph features along the random walks terminating at each node. Relating the rich literature on RNN behavior and graph topology, we theoretically show and experimentally verify that RUM attenuates the aforementioned symptoms and is more expressive than the Weisfeiler-Lehman (WL) isomorphism test. On a variety of node- and graph-level classification and regression tasks, RUM not only achieves competitive performance, but is also robust, memory-efficient, scalable, and faster than the simplest convolutional GNNs.