Relational concepts are indeed foundational to the structure of knowledge representation, as they facilitate the association between various entity concepts, allowing us to express and comprehend complex world knowledge.By expressing relational concepts in natural language prompts, people can effortlessly interact with large language models (LLMs) and recall desired factual knowledge. However, the process of knowledge recall lacks interpretability, and representations of relational concepts within LLMs remain unknown to us. In this paper, we identify hidden states that can express entity and relational concepts through causal mediation analysis in fact recall processes. Our finding reveals that at the last token position of the input prompt, there are hidden states that solely express the causal effects of relational concepts. Based on this finding, we assume that these hidden states can be treated as relational representations and we can successfully extract them from LLMs. The experimental results demonstrate high credibility of the relational representations: they can be flexibly transplanted into other fact recall processes, and can also be used as robust entity connectors. Moreover, we also show that the relational representations exhibit significant potential for controllable fact recall through relation rewriting.

Generative models, widely utilized in various applications, can often struggle with prompts corresponding to partial tokens. This struggle stems from tokenization, where partial tokens fall out of distribution during inference, leading to incorrect or nonsensical outputs. This paper examines a technique to alleviate the tokenization artifact on text completion in generative models, maintaining performance even in regular non-subword cases. The method, termed token alignment, involves backtracking to the last complete tokens and ensuring the model’s generation aligns with the prompt. This approach showcases marked improvement across many partial token scenarios, including nuanced cases like space-prefix and partial indentation, with only a minor time increase. The technique and analysis detailed in this paper contribute to the continuous advancement of generative models in handling partial inputs, bearing relevance for applications like code completion and text.

Code generation models are not robust to small perturbations, which often lead to incorrect generations and significantly degrade the performance of these models. Although improving the robustness of code generation models is crucial to enhancing user experience in real-world applications, existing research efforts do not address this issue. To fill this gap, we propose CodeFort, a framework to improve the robustness of code generation models, generalizing a large variety of code perturbations to enrich the training data and enabling various robust training strategies, mixing data augmentation, batch augmentation, adversarial logits pairing, and contrastive learning, all carefully designed to support high-throughput training. Extensive evaluations show that we increase the average robust pass rates of baseline CodeGen models from 14.79 to 21.74. We notably decrease the robustness drop rate from 95.02% to 54.95% against code-syntax perturbations.

While pre-trained language models (LM) for code have achieved great success in code completion, they generate code conditioned only on the contents within the file, i.e., in-file context, but ignore the rich semantics in other files within the same project, i.e., project-level cross-file context, a critical source of information that is especially useful in modern modular software development. Such overlooking constrains code LMs’ capacity in code completion, leading to unexpected behaviors such as generating hallucinated class member functions or function calls with unexpected arguments. In this work, we propose CoCoMIC, a novel framework that jointly learns the in-file and cross-file context on top of code LMs. To empower CoCoMIC, we develop CCFinder, a static-analysis-based tool that locates and retrieves the most relevant project-level cross-file context for code completion. CoCoMIC successfully improves the existing code LM with a 33.94% relative increase in exact match and 28.69% in identifier matching for code completion when the cross-file context is provided. Finally, we perform a series of ablation studies and share valuable insights for future research on integrating cross-file context into code LMs.

Despite exciting progress in causal language models, the expressiveness of their representations is largely limited due to poor discrimination ability. To remedy this issue, we present CONTRACLM, a novel contrastive learning framework at both the token-level and the sequence-level. We assess CONTRACLM on a variety of downstream tasks. We show that CONTRACLM enhances the discrimination of representations and bridges the gap with encoder-only models, which makes causal language models better suited for tasks beyond language generation. Specifically, we attain 44% relative improvement on the Semantic Textual Similarity tasks and 34% on Code-to-Code Search tasks. Furthermore, by improving the expressiveness of representations, CONTRACLM also boosts the source code generation capability with 9% relative improvement on execution accuracy on the HumanEval benchmark.

Code generation models have achieved impressive performance. However, they tend to be brittle as slight edits to a prompt could lead to very different generations; these robustness properties, critical for user experience when deployed in real-life applications, are not well understood. Most existing works on robustness in text or code tasks have focused on classification, while robustness in generation tasks is an uncharted area and to date there is no comprehensive benchmark for robustness in code generation. In this paper, we propose ReCode, a comprehensive robustness evaluation benchmark for code generation models. We customize over 30 transformations specifically for code on docstrings, function and variable names, code syntax, and code format. They are carefully designed to be natural in real-life coding practice, preserve the original semantic meaning, and thus provide multifaceted assessments of a model’s robustness performance. With human annotators, we verified that over 90% of the perturbed prompts do not alter the semantic meaning of the original prompt. In addition, we define robustness metrics for code generation models considering the worst-case behavior under each type of perturbation, taking advantage of the fact that executing the generated code can serve as objective evaluation. We demonstrate ReCode on SOTA models using HumanEval, MBPP, as well as function completion tasks derived from them. Interesting observations include: better robustness for CodeGen over InCoder and GPT-J; models are most sensitive to syntax perturbations; more challenging robustness evaluation on MBPP over HumanEval.

Large language models trained on code have shown great potential to increase productivity of software developers. Several execution-based benchmarks have been proposed to evaluate functional correctness of model-generated code on simple programming problems. Nevertheless, it is expensive to perform the same evaluation on complex real-world projects considering the execution cost. On the other hand, static analysis tools such as linters, which can detect errors without running the program, haven’t been well explored for evaluating code generation models. In this work, we propose a static evaluation framework to quantify static errors in Python code completions, by leveraging Abstract Syntax Trees. Compared with execution-based evaluation, our method is not only more efficient, but also applicable to code in the wild. For experiments, we collect code context from open source repos to generate one million function bodies using public models. Our static analysis reveals that Undefined Name and Unused Variable are the most common errors among others made by language models. Through extensive studies, we also show the impact of sampling temperature, model size, and context on static errors in code completions.

Data augmentation is an important method for evaluating the robustness of and enhancing the diversity of training data for natural language processing (NLP) models. In this paper, we present NL-Augmenter, a new participatory Python-based natural language (NL) augmentation framework which supports the creation of transformations (modifications to the data) and filters (data splits according to specific features). We describe the framework and an initial set of 117 transformations and 23 filters for a variety of NL tasks annotated with noisy descriptive tags. The transformations incorporate noise, intentional and accidental human mistakes, socio-linguistic variation, semantically-valid style, syntax changes, as well as artificial constructs that are unambiguous to humans. We demonstrate the efficacy of NL-Augmenter by using its transformations to analyze the robustness of popular language models. We find different models to be differently challenged on different tasks, with quasi-systematic score decreases. The infrastructure, datacards, and robustness evaluation results are publicly available on GitHub for the benefit of researchers working on paraphrase generation, robustness analysis, and low-resource NLP.

People frequently interact with information retrieval (IR) systems, however, IR models exhibit biases and discrimination towards various demographics. The in-processing fair ranking methods provides a trade-offs between accuracy and fairness through adding a fairness-related regularization term in the loss function. However, there haven’t been intuitive objective functions that depend on the click probability and user engagement to directly optimize towards this. In this work, we propose the In-Batch Balancing Regularization (IBBR) to mitigate the ranking disparity among subgroups. In particular, we develop a differentiable normed Pairwise Ranking Fairness (nPRF) and leverage the T-statistics on top of nPRF over subgroups as a regularization to improve fairness. Empirical results with the BERT-based neural rankers on the MS MARCO Passage Retrieval dataset with the human-annotated non-gendered queries benchmark (CITATION) show that our IBBR method with nPRF achieves significantly less bias with minimal degradation in ranking performance compared with the baseline.

Large-scale pre-trained sequence-to-sequence models like BART and T5 achieve state-of-the-art performance on many generative NLP tasks. However, such models pose a great challenge in resource-constrained scenarios owing to their large memory requirements and high latency. To alleviate this issue, we propose to jointly distill and quantize the model, where knowledge is transferred from the full-precision teacher model to the quantized and distilled low-precision student model. Empirical analyses show that, despite the challenging nature of generative tasks, we were able to achieve a 16.5x model footprint compression ratio with little performance drop relative to the full-precision counterparts on multiple summarization and QA datasets. We further pushed the limit of compression ratio to 27.7x and presented the performance-efficiency trade-off for generative tasks using pre-trained models. To the best of our knowledge, this is the first work aiming to effectively distill and quantize sequence-to-sequence pre-trained models for language generation tasks.

Crime reporting is a prevalent form of journalism with the power to shape public perceptions and social policies. How does the language of these reports act on readers? We seek to address this question with the SuspectGuilt Corpus of annotated crime stories from English-language newspapers in the U.S. For SuspectGuilt, annotators read short crime articles and provided text-level ratings concerning the guilt of the main suspect as well as span-level annotations indicating which parts of the story they felt most influenced their ratings. SuspectGuilt thus provides a rich picture of how linguistic choices affect subjective guilt judgments. We use SuspectGuilt to train and assess predictive models which validate the usefulness of the corpus, and show that these models benefit from genre pretraining and joint supervision from the text-level ratings and span-level annotations. Such models might be used as tools for understanding the societal effects of crime reporting.

It is challenging for current one-step retrieve-and-read question answering (QA) systems to answer questions like “Which novel by the author of ‘Armada’ will be adapted as a feature film by Steven Spielberg?” because the question seldom contains retrievable clues about the missing entity (here, the author). Answering such a question requires multi-hop reasoning where one must gather information about the missing entity (or facts) to proceed with further reasoning. We present GoldEn (Gold Entity) Retriever, which iterates between reading context and retrieving more supporting documents to answer open-domain multi-hop questions. Instead of using opaque and computationally expensive neural retrieval models, GoldEn Retriever generates natural language search queries given the question and available context, and leverages off-the-shelf information retrieval systems to query for missing entities. This allows GoldEn Retriever to scale up efficiently for open-domain multi-hop reasoning while maintaining interpretability. We evaluate GoldEn Retriever on the recently proposed open-domain multi-hop QA dataset, HotpotQA, and demonstrate that it outperforms the best previously published model despite not using pretrained language models such as BERT.

Condescending language use is caustic; it can bring dialogues to an end and bifurcate communities. Thus, systems for condescension detection could have a large positive impact. A challenge here is that condescension is often impossible to detect from isolated utterances, as it depends on the discourse and social context. To address this, we present TalkDown, a new labeled dataset of condescending linguistic acts in context. We show that extending a language-only model with representations of the discourse improves performance, and we motivate techniques for dealing with the low rates of condescension overall. We also use our model to estimate condescension rates in various online communities and relate these differences to differing community norms.

People use online platforms to seek out support for their informational and emotional needs. Here, we ask what effect does revealing one’s gender have on receiving support. To answer this, we create (i) a new dataset and method for identifying supportive replies and (ii) new methods for inferring gender from text and name. We apply these methods to create a new massive corpus of 102M online interactions with gender-labeled users, each rated by degree of supportiveness. Our analysis shows wide-spread and consistent disparity in support: identifying as a woman is associated with higher rates of support - but also higher rates of disparagement.