Publications
-
Feedback Friction: LLMs Struggle to Fully Incorporate External Feedback
Recent studies have shown LLMs possess some ability to improve their responses when given external feedback. However, it remains unclear how effectively and thoroughly these models can incorporate extrinsic feedback. In an ideal scenario, if LLMs receive near-perfect and complete feedback, we would expect them to fully integrate the feedback and change their incorrect answers to correct ones. In this paper, we systematically investigate LLMs' ability to incorporate feedback by designing a controlled experimental environment. For each problem, a solver model attempts a solution, then a feedback generator with access to near-complete ground-truth answers produces targeted feedback, after which the solver tries again. We evaluate this pipeline across a diverse range of tasks, including math reasoning, knowledge reasoning, scientific reasoning, and general multi-domain evaluations with state-of-the-art language models including Claude 3.7 (with and without extended thinking). Surprisingly, even under these near-ideal conditions, solver models consistently show resistance to feedback, a limitation that we term FEEDBACK FRICTION. To mitigate this limitation, we experiment with sampling-based strategies like progressive temperature increases and explicit rejection of previously attempted incorrect answers, which yield improvements but still fail to help models achieve target performance. We also perform a rigorous exploration of potential causes of FEEDBACK FRICTION, ruling out factors such as model overconfidence and data familiarity. We hope that highlighting this issue in LLMs and ruling out several apparent causes will help future research in self-improvement.
Dongwei Jiang , Alvin Zhang , Andrew Wang , Nicholas Andrews , Daniel Khashabi
39th Conference on Neural Information Processing Systems (NeurIPS), 2025
-
Rapidly Adapting to New Voice Spoofing: Few-Shot Detection of Synthesized Speech Under Distribution Shifts
We address the challenge of detecting synthesized speech under distribution shifts—arising from unseen synthesis methods, speakers, languages, or audio conditions—relative to the training data. Few-shot learning methods are a promising way to tackle distribution shifts by rapidly adapting on the basis of a few in-distribution samples. We propose a self-attentive prototypical network to enable more robust few-shot adaptation. To evaluate our approach, we systematically compare the performance of traditional zero-shot detectors and the proposed few-shot detectors, carefully controlling training conditions to introduce distribution shifts at evaluation time. In conditions where distribution shifts hamper the zero-shot performance, our proposed few-shot adaptation technique can quickly adapt using as few as 10 in-distribution samples—achieving upto 32% relative EER reduction on deepfakes in Japanese language and 20% relative reduction on ASVspoof 2021 Deepfake dataset.
Ashi Garg , Zexin Cai , Henry Li Xinyuan , Leibny Paola García-Perera , Kevin Duh , Sanjeev Khudanpur , Matthew Wiesner , Nicholas Andrews
IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), 2025
-
Scalable Controllable Accented TTS
We tackle the challenge of scaling accented TTS systems, expanding their capabilities to include much larger amounts of training data and a wider variety of accent labels, even for accents that are poorly represented or unlabeled in traditional TTS datasets. To achieve this, we employ two strategies: 1. Accent label discovery via a speech geolocation model, which automatically infers accent labels from raw speech data without relying solely on human annotation; 2. Timbre augmentation through kNN voice conversion to increase data diversity and model robustness. These strategies are validated on CommonVoice, where we fine-tune XTTS-v2 for accented TTS with accent labels discovered or enhanced using geolocation. We demonstrate that the resulting accented TTS model not only outperforms XTTS-v2 fine-tuned on self-reported accent labels in CommonVoice, but also existing accented TTS benchmarks.
Henry Li Xinyuan , Zexin Cai , Ashi Garg , Kevin Duh , Leibny Paola García-Perera , Sanjeev Khudanpur , Nicholas Andrews , Matthew Wiesner
IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), 2025
-
Hell or High Water: Can Language Model Agents Formulate Backup Plans?
As language model agents are applied to real world problems of increasing complexity, they will be expected to formulate plans across large search spaces. If those plans fail for reasons beyond their control, how well do language agents search for alternative ways to achieve their goals? To answer this question, we devise a benchmark where each problem has at least two ways of solving it via distinct combinations of function calls. The agent interacts with this environment by searching for relevant functions from a set over four thousand possibilities. When we disable a function the agent is calling and communicate an error to that agent via natural language, we expect it to find backup solution through trial and error. Overall, we find that language agents struggle to formulate and execute backup plans in response to environment feedback. While state-of-the-art models are often able to identify the correct function to use in the right context, they struggle to adapt to feedback from the environment and often fail to pursue alternate courses of action, even when the search space is artificially restricted. We provide a systematic analysis of the failures of both open-source and commercial models, examining the effects of search space size, as well as the benefits of scaling model size in our setting. Our analysis identifies key challenges for current generation models as well as promising directions for future work.
Andrew Wang , Sophia Hager , Adi Asija , Daniel Khashabi , Nicholas Andrews
Second Conference on Language Modeling (COLM), 2025
-
Learning Extrapolative Sequence Transformations from Markov Chains
Most successful applications of deep learning involve similar training and test conditions. However, tasks such as biological sequence design involve searching for sequences that improve desirable properties beyond previously known values, which requires novel hypotheses that \emph{extrapolate} beyond training data. In these settings, extrapolation may be achieved by using random search methods such as Markov chain Monte Carlo (MCMC), which, given an initial state, sample local transformations to approximate a target density that rewards states with the desired properties. However, even with a well-designed proposal, MCMC may struggle to explore large structured state spaces efficiently. Rather than relying on stochastic search, it would be desirable to have a model that greedily optimizes the properties of interest, successfully extrapolating in as few steps as possible. We propose to learn such a model from the Markov chains resulting from MCMC search. Specifically, our approach uses selected states from Markov chains as a source of training data for an autoregressive model, which is then able to efficiently generate novel sequences that extrapolate along the sequence-level properties of interest. The proposed approach is validated on three problems: protein sequence design, text sentiment control, and text anonymization. We find that the autoregressive model can extrapolate as well or better than MCMC, but with the additional benefits of scalability and significantly higher sample efficiency.
Sophia Hager , Aleem Khan , Andrew Wang , Nicholas Andrews
Forty-Second International Conference on Machine Learning (ICML), 2025
-
Are Paraphrases Generated by Large Language Models Invertible?
High-quality paraphrases are easy to produce using instruction-tuned language models or specialized paraphrasing models. Although this capability has a variety of benign applications, paraphrasing attacks—paraphrases applied to machine-generated texts—are known to significantly degrade the performance of machine-text detectors. This motivates us to consider the novel problem of paraphrase inversion, where, given paraphrased text, the objective is to recover an approximation of the original text. The closer the approximation is to the original text, the better machine-text detectors will perform. We propose an approach which frames the problem as translation from paraphrased text back to the original text, which requires examples of texts and corresponding paraphrases to train the inversion model. Fortunately, such training data can easily be generated, given a corpus of original texts and one or more paraphrasing models. We find that language models such as GPT-4 and Llama-3 exhibit biases when paraphrasing which an inversion model can learn with a modest amount of data. Perhaps surprisingly, we also find that such models generalize well, including to paraphrase models unseen at training time. Finally, we show that when combined with a paraphrased-text detector, our inversion models provide an effective defense against paraphrasing attacks, and overall our approach yields an average improvement of +22% AUROC across seven machine-text detectors and three different domains.
Rafael Rivera-Soto , Barry Chen , Nicholas Andrews
Findings of the ACL, 2025
-
GenVC: Self-Supervised Zero-Shot Voice Conversion
Most current zero-shot voice conversion methods rely on externally supervised components, particularly speaker encoders, for training. To explore alternatives that eliminate this dependency, this paper introduces GenVC, a novel framework that disentangles speaker identity and linguistic content from speech signals in a self-supervised manner. GenVC leverages speech tokenizers and an autoregressive, Transformer-based language model as its backbone for speech generation. This design supports large-scale training while enhancing both source speaker privacy protection and target speaker cloning fidelity. Experimental results demonstrate that GenVC achieves notably higher speaker similarity, with naturalness on par with leading zero-shot approaches. Moreover, due to its autoregressive formulation, GenVC introduces flexibility in temporal alignment, reducing the preservation of source prosody and speaker-specific traits, and making it highly effective for voice anonymization.
Zexin Cai , Henry Li Xinyuan , Ashi Garg , Leibny Paola García-Perera , Kevin Duh , Sanjeev Khudanpur , Matthew Wiesner , Nicholas Andrews
IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), 2025
-
The Impact of Automatic Speech Transcription on Speaker Attribution
Speaker attribution from speech transcripts is the task of identifying a speaker from the transcript of their speech based on patterns in their language use. This task is especially useful when the audio is unavailable (e.g. deleted) or unreliable (e.g. anonymized speech). Prior work in this area has primarily focused on the feasibility of attributing speakers using transcripts produced by human annotators. However, in real-world settings, one often only has more errorful transcripts produced by automatic speech recognition (ASR) systems. In this paper, we conduct what is, to our knowledge, the first comprehensive study of the impact of automatic transcription on speaker attribution performance. In particular, we study the extent to which speaker attribution performance degrades in the face of transcription errors, as well as how properties of the ASR system impact attribution. We find that attribution is surprisingly resilient to word-level transcription errors and that the objective of recovering the true transcript is minimally correlated with attribution performance. Overall, our findings suggest that speaker attribution on more errorful transcripts produced by ASR is as good, if not better, than attribution based on human-transcribed data, possibly because ASR transcription errors can capture speaker-specific features revealing of speaker identity.
Cristina Aggazzotti , Matthew Wiesner , Elizabeth Allyn Smith , Nicholas Andrews
Transactions of the Association for Computational Linguistics (TACL), 2025
-
Content Anonymization for Privacy in Long-form Audio
Voice anonymization techniques have been found to successfully obscure a speaker's acoustic identity in short, isolated utterances in benchmarks such as the VoicePrivacy Challenge. In practice, however, utterances seldom occur in isolation: long-form audio is commonplace in domains such as interviews, phone calls, and meetings. In these cases, many utterances from the same speaker are available, which pose a significantly greater privacy risk: given multiple utterances from the same speaker, an attacker could exploit an individual's vocabulary, syntax, and turns of phrase to re-identify them, even when their voice is completely disguised. To address this risk, we propose new content anonymization approaches. Our approach performs a contextual rewriting of the transcripts in an ASR-TTS pipeline to eliminate speaker-specific style while preserving meaning. We present results in a long-form telephone conversation setting demonstrating the effectiveness of a content-based attack on voice-anonymized speech. Then we show how the proposed content-based anonymization methods can mitigate this risk while preserving speech utility. Overall, we find that paraphrasing is an effective defense against content-based attacks and recommend that stakeholders adopt this step to ensure anonymity in long-form audio.
Cristina Aggazzotti , Ashi Garg , Zexin Cai , Nicholas Andrews
arXiv preprint arXiv:2510.12780, 2025
-
Multimodal Language Models with Modality-Specific Experts for Financial Forecasting from Interleaved Sequences of Text and Time Series
Text and time series data offer complementary views of financial markets: news articles provide narrative context about company events, while stock prices reflect how markets react to those events. However, despite their complementary nature, effectively integrating these interleaved modalities for improved forecasting remains challenging. In this work, we propose a unified neural architecture that models these interleaved sequences using modality-specific experts, allowing the model to learn unique time series patterns, while still enabling joint reasoning across modalities and preserving pretrained language understanding capabilities. To further improve multimodal understanding, we introduce a cross-modal alignment framework with a salient token weighting mechanism that learns to align representations across modalities with a focus on the most informative tokens. We demonstrate the effectiveness of our approach on a large-scale financial forecasting task, achieving state-of-the-art performance across a wide variety of strong unimodal and multimodal baselines. We develop an interpretability method that reveals insights into the value of time series-context and reinforces the design of our cross-modal alignment objective. Finally, we demonstrate that these improvements translate to meaningful economic gains in investment simulations.
Ross Koval , Nicholas Andrews , Xifeng Yan
arXiv preprint arXiv:2509.19628, 2025
-
Context-Aware Language Models for Forecasting Market Impact from Sequences of Financial News
Financial news plays a critical role in the information diffusion process in financial markets and is a known driver of stock prices. However, the information in each news article is not necessarily self-contained, often requiring a broader understanding of the historical news coverage for accurate interpretation. Further, identifying and incorporating the most relevant contextual information presents significant challenges. In this work, we explore the value of historical context in the ability of large language models to understand the market impact of financial news. We find that historical context provides a consistent and significant improvement in performance across methods and time horizons. To this end, we propose an efficient and effective contextualization method that uses a large LM to process the main article, while a small LM encodes the historical context into concise summary embeddings that are then aligned with the large model's representation space. We explore the behavior of the model through multiple qualitative and quantitative interpretability tests and reveal insights into the value of contextualization. Finally, we demonstrate that the value of historical context in model predictions has real-world applications, translating to substantial improvements in simulated investment performance.
Ross Koval , Nicholas Andrews , Xifeng Yan
arXiv preprint arXiv:2509.12519, 2025
-
Uncertainty Distillation: Teaching Language Models to Express Semantic Confidence
As large language models (LLMs) are increasingly used for factual question-answering, it becomes more important for LLMs to have the capability to communicate the likelihood that their answer is correct. For these verbalized expressions of uncertainty to be meaningful, they should reflect the error rates at the expressed level of confidence. However, when prompted to express confidence, the error rates of current LLMs are inconsistent with their communicated confidences, highlighting the need for uncertainty quantification methods. Many prior methods calculate lexical uncertainty, estimating a model's confidence in the specific string it generated. In some cases, however, it may be more useful to estimate semantic uncertainty, or the model's confidence in the answer regardless of how it is verbalized. We propose a simple procedure, uncertainty distillation, to teach an LLM to verbalize calibrated semantic confidences. Using held-out data to map initial uncertainty estimates to meaningful probabilities, we create examples annotated with verbalized probabilities for supervised fine-tuning. We compare uncertainty distillation to several strong baselines, and find that our method yields verbalized confidences that correlate well with observed error rates.
Sophia Hager , David Mueller , Kevin Duh , Nicholas Andrews
arXiv preprint arXiv:2503.14749, 2025
-
Language Models Optimized to Fool Detectors Still Have a Distinct Style (And How to Change It)
Despite considerable progress in the development of machine-text detectors, it has been suggested that the problem is inherently hard, and therefore, that stakeholders should proceed under the assumption that machine-generated text cannot be reliably detected as such. We examine a recent such claim by Nicks et al. (2024) regarding the ease with which language models can be optimized to degrade the performance of machine-text detectors, including detectors not specifically optimized against. We identify a feature space–the stylistic feature space–that is robust to such optimization, and show that it may be used to reliably detect samples from language models optimized to prevent detection. Furthermore, we show that even when models are explicitly optimized against stylistic detectors, detection performance remains surprisingly unaffected. We then seek to understand if stylistic detectors are inherently more robust. To study this question, we explore a new paraphrasing approach that simultaneously aims to close the gap between human writing and machine writing in stylistic feature space while avoiding detection using traditional features. We show that when only a single sample is available for detection, this attack is universally effective across all detectors considered, including those that use writing style. However, as the number of samples available for detection grows, the human and machine distributions become distinguishable. This observation encourages us to introduce AURA, a metric that estimates the overlap between human and machine-generated distributions by analyzing how detector performance improves as more samples become available. Overall, our findings underscore previous recommendations to avoid reliance on machine-text detection.
Rafael Rivera Soto , Barry Chen , Nicholas Andrews
arXiv preprint arXiv:2505.14608, 2025
-
ShiftySpeech: A Large-Scale Synthetic Speech Dataset with Distribution Shifts
The problem of synthetic speech detection has enjoyed considerable attention, with recent methods achieving low error rates across several established benchmarks. However, to what extent can low error rates on academic benchmarks translate to more realistic conditions? In practice, while the training set is fixed at one point in time, test-time conditions may exhibit distribution shifts relative to the training conditions, such as changes in speaker characteristics, emotional expressiveness, language and acoustic conditions, and the emergence of novel synthesis methods. Although some existing datasets target subsets of these distribution shifts, systematic analysis remains difficult due to inconsistencies between source data and synthesis systems across datasets. This difficulty is further exacerbated by the rapid development of new text-to-speech (TTS) and vocoder systems, which continually expand the diversity of synthetic speech. To enable systematic benchmarking of model performance under distribution shifts, we introduce ShiftySpeech, a large-scale benchmark comprising over 3,000 hours of synthetic speech across 7 source domains, 6 TTS systems, 12 vocoders, and 3 languages. ShiftySpeech is specifically designed to evaluate model generalization under controlled distribution shifts while ensuring broad coverage of modern synthetic speech generation techniques. It fills a key gap in current benchmarks by supporting fine-grained, controlled analysis of generalization robustness. All tested distribution shifts significantly degrade detection performance of state-of-the-art detection approaches based on self-supervised features. Overall, our findings suggest that reliance on synthetic speech detection methods in production environments should be carefully evaluated based on anticipated distribution shifts.
Ashi Garg , Zexin Cai , Lin Zhang , Henry Li Xinyuan , Leibny Paola García-Perera , Kevin Duh , Sanjeev Khudanpur , Matthew Wiesner , Nicholas Andrews
arXiv preprint arXiv:2502.05674, 2025
-
AnaloBench: Benchmarking the Identification of Abstract and Long-context Analogies
Humans regularly engage in analogical thinking, relating personal experiences to current situations (X is analogous to Y because of Z). Analogical thinking allows humans to solve problems in creative ways, grasp difficult concepts, and articulate ideas more effectively. Can language models (LMs) do the same? To answer this question, we propose AnaloBench, a benchmark to determine analogical reasoning ability in LMs. Our benchmarking approach focuses on aspects of this ability that are common among humans: (i) recalling related experiences from a large amount of information, and (ii) applying analogical reasoning to complex and lengthy scenarios. We collect a set of 340 high quality, human written analogies for use in our benchmark, which constitutes the largest such collection to date. We then test a broad collection of models consisting of 12 open source and 3 proprietary in various sizes and architectures. As in prior results, scaling up LMs results in some performance boosts. Surprisingly, scale offers minimal gains when, (i) analogies involve lengthy scenarios, or (ii) recalling relevant scenarios from a large pool of information, a process analogous to finding a needle in a haystack. We hope these observations encourage further research in this field.
Xiao Ye , Andrew Wang , Jacob Choi , Yining Lu , Shreya Sharma , Lingfeng Shen , Vijay Murari Tiyyala , Nicholas Andrews , Daniel Khashabi
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024
-
Financial Forecasting from Textual and Tabular Time Series
There is a variety of multimodal data pertinent to public companies, spanning from accounting statements, macroeconomic statistics, earnings conference calls, and financial reports. These diverse modalities capture the state of firms from a variety of different perspectives but requires complex interactions to reconcile in the formation of accurate financial predictions. The commonality between these different modalities is that they all represent a time series, typically observed for a particular firm at a quarterly horizon, providing the ability to model trends and variations of company data over time. However, the time series of these diverse modalities contains varying temporal and cross-channel patterns that are challenging to model without the appropriate inductive biases. In this work, we design a novel multimodal time series prediction task that includes numerical financial results, macroeconomic states, and long financial documents to predict next quarter's company earnings relative to analyst expectations. We explore a variety of approaches for this novel setting, establish strong unimodal baselines, and propose a multimodal model that exhibits state-of-the-art performance on this unique task. We demonstrate that each modality contains unique information and that the best performing model requires careful fusion of the different modalities in a multi-stage training approach. To better understand model behavior, we conduct a variety of probing experiments, reveal insights into the value of different modalities, and demonstrate the practical utility of our proposed method in a simulated trading setting.
Ross Koval , Nicholas Andrews , Xifeng Yan
Findings of the Association for Computational Linguistics: EMNLP 2024, 2024
-
Multi-Task Transfer Matters During Instruction-Tuning
Instruction-tuning trains a language model on hundreds of tasks jointly to improve a model’s ability to learn in-context;however, the mechanisms that drive in-context learning are poorly understood and, as a result, the role of instruction-tuning on in-context generalization is poorly understood as well.In this work, we study the impact of instruction-tuning on multi-task transfer: how well a model’s parameters adapt to an unseen task via fine-tuning.We find that instruction-tuning negatively impacts a model’s transfer to unseen tasks, and that model transfer and in-context generalization are highly correlated, suggesting that this catastrophic forgetting may impact in-context learning.We study methods to improve model transfer, finding that multi-task training—how well the training tasks are optimized—can significantly impact ICL generalization; additionally, we find that continual training on unsupervised pre-training data can mitigate forgetting and improve ICL generalization as well.Finally, we demonstrate that, early into training, the impact of instruction-tuning on model transfer to tasks impacts in-context generalization on that task.Overall, we provide significant evidence that multi-task transfer is deeply connected to a model’s ability to learn a task in-context.
David Mueller , Mark Dredze , Nicholas Andrews
Findings of the Association for Computational Linguistics ACL 2024, 2024
-
Can Optimization Trajectories Explain Multi-Task Transfer?
Despite the widespread adoption of multi-task training in deep learning, little is understood about how multi-task learning (MTL) affects generalization. Prior work has conjectured that the negative effects of MTL are due to optimization challenges that arise during training, and many optimization methods have been proposed to improve multi-task performance. However, recent work has shown that these methods fail to consistently improve multi-task generalization. In this work, we seek to improve our understanding of these failures by empirically studying how MTL impacts the optimization of tasks, and whether this impact can explain the effects of MTL on generalization. We show that MTL results in a generalization gap (a gap in generalization at comparable training loss) between single-task and multi-task trajectories early into training. However, we find that factors of the optimization trajectory previously proposed to explain generalization gaps in single-task settings cannot explain the generalization gaps between single-task and multi-task models. Moreover, we show that the amount of gradient conflict between tasks is correlated with negative effects to task optimization, but is not predictive of generalization. Our work sheds light on the underlying causes for failures in MTL and, importantly, raises questions about the role of general purpose multi-task optimization algorithms.
David Mueller , Mark Dredze , Nicholas Andrews
Transactions on Machine Learning Research (TMLR), 2024
-
Privacy Versus Emotion Preservation Trade-Offs in Emotion-Preserving Speaker Anonymization
Advances in speech technology now allow unprecedented access to personally identifiable information through speech. To protect such information, the differential privacy field has explored ways to anonymize speech while preserving its utility, including linguistic and paralinguistic aspects. However, anonymizing speech while maintaining emotional state remains challenging. We explore this problem in the context of the VoicePrivacy 2024 challenge. Specifically, we developed various speaker anonymization pipelines and find that approaches either excel at anonymization or preserving emotion state, but not both simultaneously. Achieving both would require an in-domain emotion recognizer. Additionally, we found that it is feasible to train a semi-effective speaker verification system using only emotion representations, demonstrating the challenge of separating these two modalities.
Zexin Cai , Henry Li Xinyuan , Ashi Garg , Leibny Paola Garcia-Perera , Kevin Duh , Sanjeev Khudanpur , Nicholas Andrews , Matthew Wiesner
2024 IEEE Spoken Language Technology Workshop (SLT), 2024
-
HLTCOE Submission to the 2024 Voice Privacy Challenge
We present a number of systems for the Voice Privacy Challenge, including voice conversion based systems such as the kNN-VC method and the WavLM voice Conversion method, and text-to-speech (TTS) based systems including Whisper-VITS. We found that while voice conversion systems better preserve emotional content, they struggle to conceal speaker identity in semi-white-box attack scenarios; conversely, TTS methods perform better at anonymization and worse at emotion preservation. Finally, we propose a random admixture system which seeks to balance out the strengths and weaknesses of the two category of systems, achieving a strong EER of over 40% while maintaining UAR at a respectable 47%.
Henry Li Xinyuan , Zexin Cai , Ashi Garg , Leibny Paola Garcia-Perera , Kevin Duh , Sanjeev Khudanpur , Nicholas Andrews , Matthew Wiesner
Proc. 4th Symposium on Security and Privacy in Speech Communication, 2024
Awards: Best paper
-
Can Authorship Attribution Models Distinguish Speakers in Speech Transcripts?
Authorship verification is the task of determining if two distinct writing samples share the same author and is typically concerned with the attribution of written text. In this paper, we explore the attribution of transcribed speech, which poses novel challenges. The main challenge is that many stylistic features, such as punctuation and capitalization, are not informative in this setting. On the other hand, transcribed speech exhibits other patterns, such as filler words and backchannels (e.g., um, uh-huh), which may be characteristic of different speakers. We propose a new benchmark for speaker attribution focused on human-transcribed conversational speech transcripts. To limit spurious associations of speakers with topic, we employ both conversation prompts and speakers participating in the same conversation to construct verification trials of varying difficulties. We establish the state of the art on this new benchmark by comparing a suite of neural and non-neural baselines, finding that although written text attribution models achieve surprisingly good performance in certain settings, they perform markedly worse as conversational topic is increasingly controlled. We present an analysis of the impact of transcription style on performance as well as the ability of fine-tuning on speech transcripts to improve performance.
Cristina Aggazzotti , Nicholas Andrews , Elizabeth Allyn Smith
Transactions of the Association for Computational Linguistics, 2024
-
Few-Shot Detection of Machine-Generated Text using Style Representations
The advent of instruction-tuned language models that convincingly mimic human writing poses a significant risk of abuse. However, such abuse may be counteracted with the ability to detect whether a piece of text was composed by a language model rather than a human author. Some previous approaches to this problem have relied on supervised methods by training on corpora of confirmed human- and machine- written documents. Unfortunately, model under-specification poses an unavoidable challenge for neural network-based detectors, making them brittle in the face of data shifts, such as the release of newer language models producing still more fluent text than the models used to train the detectors. Other approaches require access to the models that may have generated a document in question, which is often impractical. In light of these challenges, we pursue a fundamentally different approach not relying on samples from language models of concern at training time. Instead, we propose to leverage representations of writing style estimated from human-authored text. Indeed, we find that features effective at distinguishing among human authors are also effective at distinguishing human from machine authors, including state-of-the-art large language models like Llama-2, ChatGPT, and GPT-4. Furthermore, given a handful of examples composed by each of several specific language models of interest, our approach affords the ability to predict which model generated a given document. The code and data to reproduce our experiments are available at this https URL.
Rafael Rivera-Soto , Kailin Koch , Aleem Khan , Barry Chen , Marcus Bishop , Nicholas Andrews
International Conference on Learning Representations (ICLR), 2024
-
Learning to Compare Financial Reports for Financial Forecasting
Public companies in the US are required to publish annual reports that detail their recent financial performance, present the current state of ongoing business operations, and discuss future prospects. However, they typically contain over 25,000 words across all sections, large amounts of industry and legal jargon, and a high percentage of boilerplate content that does not change much year-to-year. These unique characteristics present challenges for many generic pretrained language models because it is likely that only a small percentage of the long report that reflects salient information contains meaningful signal about the future prospects of the company. In this work, we curate a large-scale dataset of paired financial reports and introduce two novel, challenging tasks of predicting long-horizon company risk and correlation that evaluate the ability of the model to recognize cross-document relationships with complex, nuanced signals. We explore and present a comprehensive set of methods and experiments, and establish strong baselines designed to learn to identify subtle similarities and differences between long documents. Furthermore, we demonstrate that it is possible to predict company risk and correlation solely from the text of their financial reports and further that modeling the cross-document interactions at a fine-grained level provides significant benefit. Finally, we probe the best performing model through quantitative and qualitative interpretability methods to reveal some insight into the underlying task signal.
Ross Koval , Nicholas Andrews , Xifeng Yan
Findings of the Association for Computational Linguistics: EACL, 2024
-
Forecasting Earnings Surprises from Conference Call Transcripts
There is a multitude of textual data relevant to the financial markets, spanning genres such as financial news, earnings conference calls, and social media posts. Earnings conference calls are one of the most important to information flow as they reflect a direct communication between company executives, financial analysts, and large shareholders. Since these calls contain content that is forward-looking in nature, they can be used to forecast the future performance of the company relative to market expectations. However, they typically contain over 5,000 words of text and large amounts of industry jargon. This length and domain-specific language present problems for many generic pretrained language models. In this work, we introduce a novel task of predicting earnings surprises from earnings call transcripts and contribute a new long document dataset that tests financial understanding with complex signals. We explore a variety of approaches for this long document classification task and establish some strong baselines. Furthermore, we demonstrate that it is possible to predict companies’ future earnings surprises from solely the text of their conference calls with reasonable accuracy. Finally, we probe the models through different interpretability methods and reveal some intuitive explanations of the linguistic features captured that go beyond traditional sentiment analysis.
Ross Koval , Nicholas Andrews , Xifeng Yan
Findings of the Association for Computational Linguistics: ACL 2023, 2023
-
Can Authorship Representation Learning Capture Stylistic Features?
Automatically disentangling an author’s style from the content of their writing is a longstanding and possibly insurmountable problem in computational linguistics. At the same time, the availability of large text corpora furnished with author labels has recently enabled learning authorship representations in a purely data-driven manner for authorship attribution, a task that ostensibly depends to a greater extent on encoding writing style than encoding content. However, success on this surrogate task does not ensure that such representations capture writing style since authorship could also be correlated with other latent variables, such as topic. In an effort to better understand the nature of the information these representations convey, and specifically to validate the hypothesis that they chiefly encode writing style, we systematically probe these representations through a series of targeted experiments. The results of these experiments suggest that representations learned for the surrogate authorship prediction task are indeed sensitive to writing style. As a consequence, authorship representations may be expected to be robust to certain kinds of data shift, such as topic drift over time. Additionally, our findings may open the door to downstream applications that require stylistic representations, such as style transfer.
Andrew Wang , Cristina Aggazzotti , Rebecca Kotula , Rafael Rivera Soto , Marcus Bishop , Nicholas Andrews
Transactions of the Association for Computational Linguistics, 2023
-
Learning to Generate Text in Arbitrary Writing Styles
Prior work in style-controlled text generation has focused on tasks such as emulating the style of prolific literary authors, producing formal or informal text, and mitigating toxicity of generated text. Plentiful demonstrations of these styles are available, and as a result modern language models are often able to emulate them, either via prompting or discriminative control. However, in applications such as writing assistants, it is desirable for language models to produce text in an author-specific style on the basis of a potentially small writing sample. For example, someone writing in a particular dialect may prefer writing suggestions that retain the same dialect. We find that instruction-tuned language models can struggle to reproduce author-specific style demonstrated in a prompt. Instead, we propose to guide a language model to generate text in a target style using contrastively-trained representations that capture stylometric features. Our approach (StyleMC) combines an author-adapted language model with sequence-level inference to improve stylistic consistency, and is found to be effective in a variety of conditions, including unconditional generation and style transfer. Additionally, we find that the proposed approach can serve as an effective anonymization method, by editing a document to mask authorship while preserving the original meaning.
Aleem Khan , Andrew Wang , Sophia Hager , Nicholas Andrews
arXiv preprint arXiv:2312.17242, 2023
-
Do Text-to-Text Multi-Task Learners Suffer from Task Conflict?
Traditional multi-task learning architectures learn a single model across multiple tasks through a shared encoder followed by task-specific decoders. Learning these models often requires specialized training algorithms that address task-conflict in the shared parameter updates, which otherwise can lead to negative transfer. A new type of multi-task learning within NLP homogenizes multi-task architectures as a shared encoder and language model decoder, which does surprisingly well across a range of diverse tasks. Does this new architecture suffer from task-conflicts that require specialized training algorithms? We study how certain factors in the shift towards text-to-text models affects multi-task conflict and negative transfer, finding that both directional conflict and transfer are surprisingly constant across architectures.
David Mueller , Nicholas Andrews , Mark Dredze
Findings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2022
-
The Importance of Temperature in Multi-Task Optimization
The promise of multi-task learning is that optimizing a single model on multiple related tasks will lead to a better solution for all tasks than independently trained models. In practice, optimization difficulties, such as conflicting gradients, can result in negative transfer, where multi-task models which perform worse than single-task models. In this work, we identify the optimization temperature---the ratio of learning rate to batch size---as a key factor in negative transfer. Temperature controls the level of noise in each optimization step, which prior work has shown to have a strong correlation with generalization. We demonstrate that, in some multi-task settings, negative transfer may arise due to poorly set optimization temperature, rather than inherently high task conflict. The implication of this finding is that in some settings, SGD with a carefully controlled temperature achieves comparable, and in some cases superior, performance to that of specialized optimization procedures such as PCGrad, MGDA, and GradNorm. In particular, our results suggest that the significant additional computational burden of these specialized methods may not always be necessary. Finally, we observe a conflict between the optimal temperatures of different tasks in a multi-task objective, with different levels of noise promoting better generalization for different tasks. Our work suggests the need for novel multi-task optimization methods which consider individual task noise-levels, and their impact on generalization.
David Mueller , Mark Dredze , Nicholas Andrews
OPT 2022: Optimization for Machine Learning (NeurIPS 2022 Workshop), 2022
-
Low-Resource Authorship Style Transfer: Can Non-Famous Authors Be Imitated?
Authorship style transfer involves altering text to match the style of a target author whilst preserving the original meaning. Existing unsupervised approaches like STRAP have largely focused on style transfer to target authors with many examples of their writing style in books, speeches, or other published works. This high-resource training data requirement (often greater than 100,000 words) makes these approaches primarily useful for style transfer to published authors, politicians, or other well-known figures and authorship styles, while style transfer to non-famous authors has not been well-studied. We introduce the low-resource authorship style transfer task, a more challenging class of authorship style transfer where only a limited amount of text in the target author's style may exist. In our experiments, we specifically choose source and target authors from Reddit and style transfer their Reddit posts, limiting ourselves to just 16 posts (on average ~500 words) of the target author's style. Style transfer accuracy is typically measured by how often a classifier or human judge will classify an output as written by the target author. Recent authorship representations models excel at authorship identification even with just a few writing samples, making automatic evaluation of this task possible for the first time through evaluation metrics we propose. Our results establish an in-context learning technique we develop as the strongest baseline, though we find current approaches do not yet achieve mastery of this challenging task. We release our data and implementations to encourage further investigation.
Ajay Patel , Nicholas Andrews , Chris Callison-Burch
arXiv preprint arXiv:2212.08986, 2022
-
Learning universal authorship representations
Determining whether two documents were composed by the same author, also known as authorship verification, has traditionally been tackled using statistical methods. Recently, authorship representations learned using neural networks have been found to outperform alternatives, particularly in large-scale settings involving hundreds of thousands of authors. But do such representations learned in a particular domain transfer to other domains? Or are these representations inherently entangled with domain-specific features? To study these questions, we conduct the first large-scale study of cross-domain transfer for authorship verification considering zero-shot transfers involving three disparate domains: Amazon reviews, fanfiction short stories, and Reddit comments. We find that although a surprising degree of transfer is possible between certain domains, it is not so successful between others. We examine properties of these domains that influence generalization and propose simple but effective methods to improve transfer.
Rafael A Rivera-Soto , Olivia Elizabeth Miano , Juanita Ordonez , Barry Y Chen , Aleem Khan , Marcus Bishop , Nicholas Andrews
Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, 2021
-
A Deep Metric Learning Approach to Account Linking
We consider the task of linking social media accounts that belong to the same author in an automated fashion on the basis of the content and metadata of their corresponding document streams. We focus on learning an embedding that maps variable-sized samples of user activity -- ranging from single posts to entire months of activity -- to a vector space, where samples by the same author map to nearby points. The approach does not require human-annotated data for training purposes, which allows us to leverage large amounts of social media content. The proposed model outperforms several competitive baselines under a novel evaluation framework modeled after established recognition benchmarks in other domains. Our method achieves high linking accuracy, even with small samples from accounts not seen at training time, a prerequisite for practical applications of the proposed linking framework.
Aleem Khan , Elizabeth Fleming , Noah Schofield , Marcus Bishop , Nicholas Andrews
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2021
-
Ensemble Distillation for Structured Prediction: Calibrated, Accurate, Fast—Choose Three
Modern neural networks do not always produce well-calibrated predictions, even when trained with a proper scoring function such as cross-entropy. In classification settings, simple methods such as isotonic regression or temperature scaling may be used in conjunction with a held-out dataset to calibrate model outputs. However, extending these methods to structured prediction is not always straightforward or effective; furthermore, a held-out calibration set may not always be available. In this paper, we study ensemble distillation as a general framework for producing well-calibrated structured prediction models while avoiding the prohibitive inference-time cost of ensembles. We validate this framework on two tasks: named-entity recognition and machine translation. We find that, across both tasks, ensemble distillation produces models which retain much of, and occasionally improve upon, the performance and calibration benefits of ensembles, while only requiring a single model during test-time.
Steven Reich , David Mueller , Nicholas Andrews
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2020
-
Sources of Transfer in Multilingual Named Entity Recognition
Named-entities are inherently multilingual, and annotations in any given language may be limited. This motivates us to consider polyglot named-entity recognition (NER), where one model is trained using annotated data drawn from more than one language. However, a straightforward implementation of this simple idea does not always work in practice: naive training of NER models using annotated data drawn from multiple languages consistently underperforms models trained on monolingual data alone, despite having access to more training data. The starting point of this paper is a simple solution to this problem, in which polyglot models are fine-tuned on monolingual data to consistently and significantly outperform their monolingual counterparts. To explain this phenomena, we explore the sources of multilingual transfer in polyglot NER models and examine the weight structure of polyglot models compared to their monolingual counterparts. We find that polyglot models efficiently share many parameters across languages and that fine-tuning may utilize a large number of those parameters.
David Mueller , Nicholas Andrews , Mark Dredze
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 2020
-
Compressing BERT: Studying the Effects of Weight Pruning on Transfer Learning
Pre-trained universal feature extractors, such as BERT for natural language processing and VGG for computer vision, have become effective methods for improving deep learning models without requiring more labeled data. While effective, feature extractors like BERT may be prohibitively large for some deployment scenarios. We explore weight pruning for BERT and ask: how does compression during pre-training affect transfer learning? We find that pruning affects transfer learning in three broad regimes. Low levels of pruning (30-40%) do not affect pre-training loss or transfer to downstream tasks at all. Medium levels of pruning increase the pre-training loss and prevent useful pre-training information from being transferred to downstream tasks. High levels of pruning additionally prevent models from fitting downstream datasets, leading to further degradation. Finally, we observe that fine-tuning BERT on a specific task does not improve its prunability. We conclude that BERT can be pruned once during pre-training rather than separately for each task without affecting performance.
Mitchell Gordon , Kevin Duh , Nicholas Andrews
Proceedings of the 5th Workshop on Representation Learning for NLP, 2020
-
Learning Invariant Representations of Social Media Users
The evolution of social media users’ behavior over time complicates user-level comparison tasks such as verification, classification, clustering, and ranking. As a result, naive approaches may fail to generalize to new users or even to future observations of previously known users. In this paper, we propose a novel procedure to learn a mapping from short episodes of user activity on social media to a vector space in which the distance between points captures the similarity of the corresponding users’ invariant features. We fit the model by optimizing a surrogate metric learning objective over a large corpus of unlabeled social media content. Once learned, the mapping may be applied to users not seen at training time and enables efficient comparisons of users in the resulting vector space. We present a comprehensive evaluation to validate the benefits of the proposed approach using data from Reddit, Twitter, and Wikipedia.
Nicholas Andrews , Marcus Bishop
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 2019
-
Convolutions Are All You Need (For Classifying Character Sequences)
While recurrent neural networks (RNNs) are widely used for text classification, they demonstrate poor performance and slow convergence when trained on long sequences. When text is modeled as characters instead of words, the longer sequences make RNNs a poor choice. Convolutional neural networks (CNNs), although somewhat less ubiquitous than RNNs, have an internal structure more appropriate for long-distance character dependencies. To better understand how CNNs and RNNs differ in handling long sequences, we use them for text classification tasks in several character-level social media datasets. The CNN models vastly outperform the RNN models in our experiments, suggesting that CNNs are superior to RNNs at learning to classify character-level data.
Zach Wood-Doughty , Nicholas Andrews , Mark Dredze
Proceedings of the 2018 EMNLP Workshop W-NUT: The 4th Workshop on Noisy User-generated Text, 2018
-
Predicting Twitter User Demographics from Names Alone
Social media analysis frequently requires tools that can automatically infer demographics to contextualize trends. These tools often require hundreds of user-authored messages for each user, which may be prohibitive to obtain when analyzing millions of users. We explore character-level neural models that learn a representation of a user's name and screen name to predict gender and ethnicity, allowing for demographic inference with minimal data. We release trained models1 which may enable new demographic analyses that would otherwise require enormous amounts of data collection
Zach Wood-Doughty , Nicholas Andrews , Rebecca Marvin , Mark Dredze
Proceedings of the Second Workshop on Computational Modeling of People's Opinions, Personality, and Emotions in Social Media, 2018
-
Bayesian Modeling of Lexical Resources for Low-Resource Settings
Lexical resources such as dictionaries and gazetteers are often used as auxiliary data for tasks such as part-of-speech induction and named-entity recognition. However, discriminative training with lexical features requires annotated data to reliably estimate the lexical feature weights and may result in overfitting the lexical features at the expense of features which generalize better. In this paper, we investigate a more robust approach: we stipulate that the lexicon is the result of an assumed generative process. Practically, this means that we may treat the lexical resources as observations under the proposed generative model. The lexical resources provide training data for the generative model without requiring separate data to estimate lexical feature weights. We evaluate the proposed approach in two settings: part-of-speech induction and low-resource named-entity recognition.
Nicholas Andrews , Mark Dredze , Benjamin Van Durme , Jason Eisner
Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2017
-
Twitter at the Grammys: A Social Media Corpus for Entity Linking and Disambiguation
Work on cross document coreference resolution (CDCR) has primarily focused on news articles, with little to no work for social media. Yet social media may be particularly challenging since short messages provide little context, and informal names are pervasive. We introduce a new Twitter corpus that contains entity annotations for entity clusters that supports CDCR. Our corpus draws from Twitter data surrounding the 2013 Grammy music awards ceremony, providing a large set of annotated tweets focusing on a single event. To establish a baseline we evaluate two CDCR systems and consider the performance impact of each system component. Furthermore, we augment one system to include temporal information, which can be helpful when documents (such as tweets) arrive in a specific order. Finally, we include annotations linking the entities to a knowledge base to support entity linking. Our corpus is available: https://bitbucket.org/mdredze/tgx
Mark Dredze , Nicholas Andrews , Jay DeYoung
Proceedings of The Fourth International Workshop on Natural Language Processing for Social Media, 2016
-
Robust Entity Clustering via Phylogenetic Inference
Entity clustering must determine when two named-entity mentions refer to the same entity. Typical approaches use a pipeline architecture that clusters the mentions using fixed or learned measures of name and context similarity. In this paper, we propose a model for cross-document coreference resolution that achieves robustness by learning similarity from unlabeled data. The generative process assumes that each entity mention arises from copying and optionally mutating an earlier name from a similar context. Clustering the mentions into entities depends on recovering this copying tree jointly with estimating models of the mutation process and parent selection process. We present a block Gibbs sampler for posterior inference and an empirical evaluation on several datasets.
Nicholas Andrews , Jason Eisner , Mark Dredze
Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2014
-
Parma: A predicate argument aligner
We introduce PARMA, a system for cross-document, semantic predicate and argument alignment. Our system combines a number of linguistic resources familiar to researchers in areas such as recognizing textual entailment and question answering, integrating them into a simple discriminative model. PARMA achieves state of the art results on an existing and a new dataset. We suggest that previous efforts have focussed on data that is biased and too easy, and we provide a more difficult dataset based on translation data with a low baseline which we beat by 17% F1.
Travis Wolfe , Benjamin Van Durme , Mark Dredze , Nicholas Andrews , Charley Beller , Chris Callison-Burch , Jay DeYoung , Justin Snyder , Jonathan Weese , Tan Xu , others
Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), 2013
-
Name Phylogeny: A Generative Model of String Variation
Many linguistic and textual processes involve transduction of strings. We show how to learn a stochastic transducer from an unorganized collection of strings (rather than string pairs). The role of the transducer is to organize the collection. Our generative model explains similarities among the strings by supposing that some strings in the collection were not generated ab initio, but were instead derived by transduction from other, “similar” strings in the collection. Our variational EM learning algorithm alternately reestimates this phylogeny and the transducer parameters. The final learned transducer can quickly link any test name into the final phylogeny, thereby locating variants of the test name. We find that our method can effectively find name variants in a corpus of web strings used to refer to persons in Wikipedia, improving over standard untrained distances such as Jaro-Winkler and Levenshtein distance.
Nicholas Andrews , Jason Eisner , Mark Dredze
Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, 2012
-
Entity Clustering Across Languages
Standard entity clustering systems commonly rely on mention (string) matching, syntactic features, and linguistic resources like English WordNet. When co-referent text mentions appear in different languages, these techniques cannot be easily applied. Consequently, we develop new methods for clustering text mentions across documents and languages simultaneously, producing cross-lingual entity clusters. Our approach extends standard clustering algorithms with cross-lingual mention and context similarity measures. Crucially, we do not assume a pre-existing entity list (knowledge base), so entity characteristics are unknown. On an Arabic-English corpus that contains seven different text genres, our best model yields a 24.3% F1 gain over the baseline.
Spence Green , Nicholas Andrews , Matthew R. Gormley , Mark Dredze , Christopher D. Manning
Proceedings of the 2012 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2012
-
Transformation Process Priors
Nicholas Andrews , Jason Eisner
NeurIPS Workshop on Bayesian Nonparametrics: Hope or Hype?, 2011
-
Seeded Discovery of Base Relations in Large Corpora
Relationship discovery is the task of identifying salient relationships between named entities in text. We propose novel approaches for two sub-tasks of the problem: identifying the entities of interest, and partitioning and describing the relations based on their semantics. In particular, we show that term frequency patterns can be used effectively instead of supervised NER, and that the p-median clustering objective function naturally uncovers relation exemplars appropriate for describing the partitioning. Furthermore, we introduce a novel application of relationship discovery: the unsupervised identification of protein-protein interaction phrases.
Nicholas Andrews , Naren Ramakrishnan
Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing, 2008