Since most large language models (LLMs) are trained once and never updated, they struggle to dynamically adapt to our ever-changing world. In this work, we present FreshQA, a dynamic QA benchmark that tests a model’s ability to answer questions that may require reasoning over up-to-date world knowledge. We develop a two-mode human evaluation procedure to measure both correctness and hallucination, which we use to benchmark both closed and open-source LLMs by collecting >50K human judgments. We observe that all LLMs struggle to answer questions that require fast-changing world knowledge as well as questions with false premises that need to be debunked. In response, we develop FreshPrompt, a few-shot prompting method that curates and organizes relevant information from a search engine into an LLM’s prompt. Our experiments show that FreshPrompt outperforms both competing search engine-augmented prompting methods such as Self-Ask (Press et al., 2022) as well as commercial systems such as Perplexity.AI. To facilitate future work, we additionally develop FreshEval, a reliable autorater for quick evaluation and comparison on FreshQA. Our latest results with FreshEval suggest that open-source LLMs such as Mixtral (Jiang et al., 2024), when combined with FreshPrompt, are competitive with closed-source and commercial systems on search-augmented QA.

Pretraining data design is critically under-documented and often guided by empirically unsupported intuitions. We pretrain models on data curated (1) at different collection times, (2) with varying toxicity and quality filters, and (3) with different domain compositions. First, we find that temporal shift between evaluation data and pretraining data leads to performance degradation, which is not overcome by finetuning. Second, we measure the effect of quality and toxicity filters, showing a trade-off between performance on standard benchmarks and risk of toxic generations. We also find that the effects of different types of filtering are not predictable from text domain characteristics. Third, we empirically validate that heterogeneous data sources, like books and web, are beneficial and warrant greater prioritization. To date, these experiments constitute the single largest publicly documented empirical study of the effects of pretraining data. Spanning 28 unique 1.5 billion parameter models pretrained from scratch, these findings validate, quantify, and expose many undocumented intuitions about text pretraining, which ultimately support more informed data-centric decisions in model development.

BIG-Bench (Srivastava et al., 2022) is a diverse evaluation suite that focuses on tasks believed to be beyond the capabilities of current language models. Language models have already made good progress on this benchmark, with the best model in the BIG-Bench paper outperforming average reported human-rater results on 65% of the BIG-Bench tasks via few-shot prompting. But on what tasks do language models fall short of average human-rater performance, and are those tasks actually unsolvable by current language models? In this work, we focus on a suite of 23 challenging BIG-Bench tasks which we call BIG-Bench Hard (BBH). These are the tasks for which prior language model evaluations did not outperform the average human-rater. We find that applying chain-of-thought (CoT) prompting to BBH tasks enables PaLM to surpass the average human-rater performance on 10 of the 23 tasks, and Codex (code-davinci-002) to surpass the average human-rater performance on 17 of the 23 tasks. Since many tasks in BBH require multi-step reasoning, few-shot prompting without CoT, as done in the BIG-Bench evaluations (Srivastava et al., 2022), substantially underestimates the best performance and capabilities of language models, which is better captured via CoT prompting. As further analysis, we explore the interaction between CoT and model scale on BBH, finding that CoT enables emergent task performance on several BBH tasks with otherwise flat scaling curves.

Scaling language models improves performance but comes with significant computational costs. This paper proposes UL2R, a method that substantially improves existing language models and their scaling curves with a relatively tiny amount of extra compute. The key idea is to continue training a state-of-the-art large language model on a few more steps with UL2’s mixture-of-denoiser objective. We show that, with almost negligible extra computational costs and no new sources of data, we are able to substantially improve the scaling properties of large language models on downstream metrics. In this paper, we continue training a baseline language model, PaLM, with ULR2, introducing a new set of models at 8B, 62B, and 540B scale which we call U-PaLM. Impressively, at 540B scale, we show an approximately 2x computational savings rate where U-PaLM achieves the same performance as the final PaLM 540B model at around half its computational budget (i.e., saving ~4.4 million TPUv4 hours). We further show that this improved scaling curve leads to “emergent abilities” on challenging BIG-Bench tasks—for instance, U-PaLM does much better on some tasks or demonstrates better quality at much smaller scale (62B as opposed to 540B). Overall, we show that U-PaLM outperforms PaLM on many few-shot setups, including reasoning tasks with chain-of-thought (e.g., GSM8K), multilingual tasks (MGSM, TydiQA), MMLU and challenging BIG-Bench tasks.

Scaling up language models has been empirically shown to improve performance on a wide range of downstream tasks. However, if we were to observe worse performance as a function of scale (inverse scaling) on certain tasks, this would indicate that scaling can also encourage behaviors that are misaligned with human preferences. The Inverse Scaling Prize (McKenzie et al. 2023) identified eleven such inverse scaling tasks, evaluated on models of up to 280B parameters and up to 500 zettaFLOPs of training compute. This paper takes a closer look at these inverse scaling tasks. In this paper, we evaluate models of up to 540B parameters, trained on five times more compute than those evaluated in the Inverse Scaling Prize. With this increased range of model sizes and compute, only four out of the eleven tasks remain inverse scaling. Six tasks exhibit U-shaped scaling, where performance decreases up to a certain size, and then increases again up to the largest model evaluated (the one remaining task displays positive scaling). In addition, 1-shot examples and chain-of-thought can help mitigate undesirable scaling patterns even further. U-shaped scaling suggests that the inverse scaling trend observed in McKenzie et al. (2023) may not continue to hold for larger models, which we attribute to the presence of distractor tasks that only sufficiently large models can avoid.

In this paper, we leverage large language models (LLMs) to perform zero-shot text style transfer. We present a prompting method that we call augmented zero-shot learning, which frames style transfer as a sentence rewriting task and requires only a natural language instruction, without model fine-tuning or exemplars in the target style. Augmented zero-shot learning is simple and demonstrates promising results not just on standard style transfer tasks such as sentiment, but also on arbitrary transformations such as ‘make this melodramatic’ or ‘insert a metaphor.’

Pre-trained language models perform well on a variety of linguistic tasks that require symbolic reasoning, raising the question of whether such models implicitly represent abstract symbols and rules. We investigate this question using the case study of BERT’s performance on English subject–verb agreement. Unlike prior work, we train multiple instances of BERT from scratch, allowing us to perform a series of controlled interventions at pre-training time. We show that BERT often generalizes well to subject–verb pairs that never occurred in training, suggesting a degree of rule-governed behavior. We also find, however, that performance is heavily influenced by word frequency, with experiments showing that both the absolute frequency of a verb form, as well as the frequency relative to the alternate inflection, are causally implicated in the predictions BERT makes at inference time. Closer analysis of these frequency effects reveals that BERT’s behavior is consistent with a system that correctly applies the SVA rule in general but struggles to overcome strong training priors and to estimate agreement features (singular vs. plural) on infrequent lexical items.

This paper asks whether extrapolating the hidden space distribution of text examples from one class onto another is a valid inductive bias for data augmentation. To operationalize this question, I propose a simple data augmentation protocol called “good-enough example extrapolation” (GE3). GE3 is lightweight and has no hyperparameters. Applied to three text classification datasets for various data imbalance scenarios, GE3 improves performance more than upsampling and other hidden-space data augmentation methods.

The uniform information density (UID) hypothesis, which posits that speakers behaving optimally tend to distribute information uniformly across a linguistic signal, has gained traction in psycholinguistics as an explanation for certain syntactic, morphological, and prosodic choices. In this work, we explore whether the UID hypothesis can be operationalized as an inductive bias for statistical language modeling. Specifically, we augment the canonical MLE objective for training language models with a regularizer that encodes UID. In experiments on ten languages spanning five language families, we find that using UID regularization consistently improves perplexity in language models, having a larger effect when training data is limited. Moreover, via an analysis of generated sequences, we find that UID-regularized language models have other desirable properties, e.g., they generate text that is more lexically diverse. Our results not only suggest that UID is a reasonable inductive bias for language modeling, but also provide an alternative validation of the UID hypothesis using modern-day NLP tools.

Although automated metrics are commonly used to evaluate NLG systems, they often correlate poorly with human judgements. Newer metrics such as BERTScore have addressed many weaknesses in prior metrics such as BLEU and ROUGE, which rely on n-gram matching. These newer methods, however, are still limited in that they do not consider the generation context, so they cannot properly reward generated text that is correct but deviates from the given reference. In this paper, we propose Language Model Augmented Relevance Score (MARS), a new context-aware metric for NLG evaluation. MARS leverages off-the-shelf language models, guided by reinforcement learning, to create augmented references that consider both the generation context and available human references, which are then used as additional references to score generated text. Compared with seven existing metrics in three common NLG tasks, MARS not only achieves higher correlation with human reference judgements, but also differentiates well-formed candidates from adversarial samples to a larger degree.

The complexity loss paradox, which posits that individuals suffering from disease exhibit surprisingly predictable behavioral dynamics, has been observed in a variety of both human and animal physiological systems. The recent advent of online text-based therapy presents a new opportunity to analyze the complexity loss paradox in a novel operationalization: linguistic complexity loss in text-based therapy conversations. In this paper, we analyze linguistic complexity correlates of mental health in the online therapy messages sent between therapists and 7,170 clients who provided 30,437 corresponding survey responses on their anxiety. We found that when clients reported more anxiety, they showed reduced lexical diversity as estimated by the moving average type-token ratio. Therapists, on the other hand, used language of higher reading difficulty, syntactic complexity, and age of acquisition when clients were more anxious. Finally, we found that clients, and to an even greater extent, therapists, exhibited consistent levels of many linguistic complexity measures. These results demonstrate how linguistic analysis of text-based communication can be leveraged as a marker for anxiety, an exciting prospect in a time of both increased online communication and increased mental health issues.

Few-shot text classification is a fundamental NLP task in which a model aims to classify text into a large number of categories, given only a few training examples per category. This paper explores data augmentation—a technique particularly suitable for training with limited data—for this few-shot, highly-multiclass text classification setting. On four diverse text classification tasks, we find that common data augmentation techniques can improve the performance of triplet networks by up to 3.0% on average. To further boost performance, we present a simple training strategy called curriculum data augmentation, which leverages curriculum learning by first training on only original examples and then introducing augmented data as training progresses. We explore a two-stage and a gradual schedule, and find that, compared with standard single-stage training, curriculum data augmentation trains faster, improves performance, and remains robust to high amounts of noising from augmentation.

Traditional data augmentation aims to increase the coverage of the input distribution by generating augmented examples that strongly resemble original samples in an online fashion where augmented examples dominate training. In this paper, we propose an alternative perspective—a multi-task view (MTV) of data augmentation—in which the primary task trains on original examples and the auxiliary task trains on augmented examples. In MTV data augmentation, both original and augmented samples are weighted substantively during training, relaxing the constraint that augmented examples must resemble original data and thereby allowing us to apply stronger augmentation functions. In empirical experiments using four common data augmentation techniques on three benchmark text classification datasets, we find that using the MTV leads to higher and more robust performance than traditional augmentation.

We present COVID-Q, a set of 1,690 questions about COVID-19 from 13 sources, which we annotate into 15 question categories and 207 question clusters. The most common questions in our dataset asked about transmission, prevention, and societal effects of COVID, and we found that many questions that appeared in multiple sources were not answered by any FAQ websites of reputable organizations such as the CDC and FDA. We post our dataset publicly at https://github.com/JerryWei03/COVID-Q. For classifying questions into 15 categories, a BERT baseline scored 58.1% accuracy when trained on 20 examples per category, and for a question clustering task, a BERT + triplet loss baseline achieved 49.5% accuracy. We hope COVID-Q can help either for direct use in developing applied systems or as a domain-specific resource for model evaluation.

The field of NLP has seen unprecedented achievements in recent years. Most notably, with the advent of large-scale pre-trained Transformer-based language models, such as BERT, there has been a noticeable improvement in text representation. It is, however, unclear whether these improvements translate to noisy user-generated text, such as tweets. In this paper, we present an experimental survey of a wide range of well-known text representation techniques for the task of text clustering on noisy Twitter data. Our results indicate that the more advanced models do not necessarily work best on tweets and that more exploration in this area is needed.

We present EDA: easy data augmentation techniques for boosting performance on text classification tasks. EDA consists of four simple but powerful operations: synonym replacement, random insertion, random swap, and random deletion. On five text classification tasks, we show that EDA improves performance for both convolutional and recurrent neural networks. EDA demonstrates particularly strong results for smaller datasets; on average, across five datasets, training with EDA while using only 50% of the available training set achieved the same accuracy as normal training with all available data. We also performed extensive ablation studies and suggest parameters for practical use.