Search

Despite great progress in object detection, most existing methods work only on a limited set of object categories, due to the tremendous human effort needed for bounding-box annotations of training data. To alleviate the problem, recent open vocabulary and zero-shot detection methods attempt to detect novel object categories beyond those seen during training. They achieve this goal by training on a pre-defined base categories to induce generalization to novel objects. However, their potential is still constrained by the small set of base categories available for training. To enlarge the set of base classes, we propose a method to automatically generate pseudo bounding-box annotations of diverse objects from large-scale image-caption pairs. Our method leverages the localization ability of pre-trained vision-language models to generate pseudo bounding-box labels and then directly uses them for training object detectors. Experimental results show that our method outperforms the state-of-the-art open vocabulary detector by 8% AP on COCO novel categories, by 6.3% AP on PASCAL VOC, by 2.3% AP on Objects365 and by 2.8% AP on LVIS. Code is available at https://github.com/salesforce/PB-OVD., Comment: ECCV 2022

Fact-checking is an essential tool to mitigate the spread of misinformation and disinformation. We introduce the task of fact-checking in dialogue, which is a relatively unexplored area. We construct DialFact, a testing benchmark dataset of 22,245 annotated conversational claims, paired with pieces of evidence from Wikipedia. There are three sub-tasks in DialFact: 1) Verifiable claim detection task distinguishes whether a response carries verifiable factual information; 2) Evidence retrieval task retrieves the most relevant Wikipedia snippets as evidence; 3) Claim verification task predicts a dialogue response to be supported, refuted, or not enough information. We found that existing fact-checking models trained on non-dialogue data like FEVER fail to perform well on our task, and thus, we propose a simple yet data-efficient solution to effectively improve fact-checking performance in dialogue. We point out unique challenges in DialFact such as handling the colloquialisms, coreferences and retrieval ambiguities in the error analysis to shed light on future research in this direction., Comment: ACL 2022

Asking good questions is an essential ability for both human and machine intelligence. However, existing neural question generation approaches mainly focus on the short factoid type of answers. In this paper, we propose a neural question generator, MixQG, to bridge this gap. We combine 9 question answering datasets with diverse answer types, including yes/no, multiple-choice, extractive, and abstractive answers, to train a single generative model. We show with empirical results that our model outperforms existing work in both seen and unseen domains and can generate questions with different cognitive levels when conditioned on different answer types. Our code is released and well-integrated with the Huggingface library to facilitate various downstream applications., Comment: camera-ready version

Hallucination is a known issue for neural abstractive summarization models. Recent work suggests that the degree of hallucination may depend on errors in the training data. In this work, we propose a new method called Contrastive Parameter Ensembling (CaPE) to use training data more effectively, utilizing variations in noise in training samples to reduce hallucination. We first select clean and noisy subsets from the training data using different automatic factual metrics. Then, we fine-tune a base summarization model, which is trained on all training samples, on the clean (noisy) subset to obtain an \textit{expert} (\textit{anti-expert}) model. Finally, we adjust the parameters of base model by the difference between parameters of the \textit{expert} and \textit{anti-expert} models, steering the base model towards the \textit{expert} model and away from the \textit{anti-expert} model. Experimental results show that CaPE improves performance across different automatic factual metrics and human evaluation, with the maximum improvement of 16.69\% and 15.78\% on summary-level dependency-arc entailment accuracy for the XSUM and CNN/DM datasets. The improvement in factual performance does not degrade the performance on other metrics of informativeness such as ROUGE.

Existing studies addressing gender bias of pre-trained language models, usually build a small gender-neutral data set and conduct a second phase pre-training on the model with such data. However, given the limited size and concentrated focus of the gender-neutral data, catastrophic forgetting would occur during second-phase pre-training. Forgetting information in the original training data may damage the model's downstream performance by a large margin. In this work, we empirically show that catastrophic forgetting occurs in such methods by evaluating them with general NLP tasks in GLUE. Then, we propose a new method, GEnder Equality Prompt (GEEP), to improve gender fairness of pre-trained models with less forgetting. GEEP freezes the pre-trained model and learns gender-related prompts with gender-neutral data. Empirical results show that GEEP not only achieves SOTA performances on gender fairness tasks, but also forgets less and performs better on GLUE by a large margin., Comment: This paper has been accepted at the 61st Annual Meeting of the Association for Computational Linguistics (ACL 2023)

Summarization systems make numerous "decisions" about summary properties during inference, e.g. degree of copying, specificity and length of outputs, etc. However, these are implicitly encoded within model parameters and specific styles cannot be enforced. To address this, we introduce HydraSum, a new summarization architecture that extends the single decoder framework of current models to a mixture-of-experts version with multiple decoders. We show that HydraSum's multiple decoders automatically learn contrasting summary styles when trained under the standard training objective without any extra supervision. Through experiments on three summarization datasets (CNN, Newsroom and XSum), we show that HydraSum provides a simple mechanism to obtain stylistically-diverse summaries by sampling from either individual decoders or their mixtures, outperforming baseline models. Finally, we demonstrate that a small modification to the gating strategy during training can enforce an even stricter style partitioning, e.g. high- vs low-abstractiveness or high- vs low-specificity, allowing users to sample from a larger area in the generation space and vary summary styles along multiple dimensions., Comment: EMNLP2022

We report a layered ternary selenide BaPt4Se6 featuring sesqui-selenide Pt2Se3 layers sandwiched by Ba atoms. The Pt2Se3 layers in this compound can be derived from the Dirac-semimetal PtSe2 phase with Se vacancies that form a honeycomb structure. This structure results in a Pt (VI) and Pt (II) mixed-valence compound with both PtSe6 octahedra and PtSe4 square net coordination configurations. Temperature dependent electrical transport measurements suggest two distinct anomalies: a resistivity crossover, mimic to the metal-insulator (M-I) transition at ~150K, and a resistivity plateau at temperatures below 10K. The resistivity crossover is not associated with any structural, magnetic or charge order modulated phase transitions. Magnetoresistivity, Hall and heat capacity measurements concurrently suggest an existing hidden state below 5K in this system. Angle-resolved photoemission spectroscopy measurements reveal a metallic state and no dramatic reconstruction of the electronic structure up to 200K., Comment: 25 pages, 5 figures

In this paper, we aim to improve abstractive dialogue summarization quality and, at the same time, enable granularity control. Our model has two primary components and stages: 1) a two-stage generation strategy that generates a preliminary summary sketch serving as the basis for the final summary. This summary sketch provides a weakly supervised signal in the form of pseudo-labeled interrogative pronoun categories and key phrases extracted using a constituency parser. 2) A simple strategy to control the granularity of the final summary, in that our model can automatically determine or control the number of generated summary sentences for a given dialogue by predicting and highlighting different text spans from the source text. Our model achieves state-of-the-art performance on the largest dialogue summarization corpus SAMSum, with as high as 50.79 in ROUGE-L score. In addition, we conduct a case study and show competitive human evaluation results and controllability to human-annotated summaries., Comment: ACL-Findings 2021. Code is released at https://github.com/salesforce/ConvSumm

This paper introduces QAConv, a new question answering (QA) dataset that uses conversations as a knowledge source. We focus on informative conversations, including business emails, panel discussions, and work channels. Unlike open-domain and task-oriented dialogues, these conversations are usually long, complex, asynchronous, and involve strong domain knowledge. In total, we collect 34,608 QA pairs from 10,259 selected conversations with both human-written and machine-generated questions. We use a question generator and a dialogue summarizer as auxiliary tools to collect and recommend questions. The dataset has two testing scenarios: chunk mode and full mode, depending on whether the grounded partial conversation is provided or retrieved. Experimental results show that state-of-the-art pretrained QA systems have limited zero-shot performance and tend to predict our questions as unanswerable. Our dataset provides a new training and evaluation testbed to facilitate QA on conversations research., Comment: ACL 2022. Data and code are available at https://github.com/salesforce/QAConv

We report a combined experimental and theoretical study of the PdSe2-xTex system. With increasing Te fraction, structural evolutions, first from an orthorhombic phase (space group Pbca) to a monoclinic phase (space group C2/c) and then to a trigonal phase (space group P-3m1), are observed accompanied with clearly distinct electrical transport behavior. The monoclinic phase (C2/c) is a completely new polymorphic phase and is discovered within a narrow range of Te composition (0.3 \leq x \leq 0.8). This phase has a different packing sequence from all known transition metal dichalcogenides to date. Electronic calculations and detailed transport analysis of the new polymorphic PdSe1.3Te0.7 phase are presented. In the trigonal phase region, superconductivity with enhanced critical temperature is also observed within a narrow range of Te content (1.0 \leq x \leq 1.2). The rich phase diagram, new polymorphic structure as well as abnormally enhanced superconductivity could further stimulate more interest to explore new types of polymorphs and investigate their transport and electronic properties in the transition metal dichalcogenides family that are of significant interest., Comment: 16 pages, 7 figures

Two-dimensional transition metal dichalcogenide PdTe$_2$ recently attracts much attention due to its phase coexistence of type-II Dirac semimetal and type-I superconductivity. Here we report a 67 % enhancement of superconducting transition temperature in the 1T-PdSeTe in comparison to that of PdTe2 through partial substitution of Te atoms by Se. The superconductivity has been unambiguously confirmed by the magnetization, resistivity and specific heat measurements. 1T-PdSeTe shows type-II superconductivity with large anisotropy and non-bulk superconductivity nature with volume fraction ~ 20 % estimated from magnetic and heat capacity measurements. 1T-PdSeTe expands the family of superconducting transition metal dichalcogenides and thus provides additional insights for understanding superconductivity and topological physics in the 1T-PdTe$_2$ system, Comment: 7 pages, 6 figures

Fitting complex patterns in the training data, such as reasoning and commonsense, is a key challenge for language pre-training. According to recent studies and our empirical observations, one possible reason is that some easy-to-fit patterns in the training data, such as frequently co-occurring word combinations, dominate and harm pre-training, making it hard for the model to fit more complex information. We argue that mis-predictions can help locate such dominating patterns that harm language understanding. When a mis-prediction occurs, there should be frequently co-occurring patterns with the mis-predicted word fitted by the model that lead to the mis-prediction. If we can add regularization to train the model to rely less on such dominating patterns when a mis-prediction occurs and focus more on the rest more subtle patterns, more information can be efficiently fitted at pre-training. Following this motivation, we propose a new language pre-training method, Mis-Predictions as Harm Alerts (MPA). In MPA, when a mis-prediction occurs during pre-training, we use its co-occurrence information to guide several heads of the self-attention modules. Some self-attention heads in the Transformer modules are optimized to assign lower attention weights to the words in the input sentence that frequently co-occur with the mis-prediction while assigning higher weights to the other words. By doing so, the Transformer model is trained to rely less on the dominating frequently co-occurring patterns with mis-predictions while focus more on the rest more complex information when mis-predictions occur. Our experiments show that MPA expedites the pre-training of BERT and ELECTRA and improves their performances on downstream tasks., Comment: Accepted at SSL workshop at NeurIPS 2021

Intent detection is one of the core components of goal-oriented dialog systems, and detecting out-of-scope (OOS) intents is also a practically important skill. Few-shot learning is attracting much attention to mitigate data scarcity, but OOS detection becomes even more challenging. In this paper, we present a simple yet effective approach, discriminative nearest neighbor classification with deep self-attention. Unlike softmax classifiers, we leverage BERT-style pairwise encoding to train a binary classifier that estimates the best matched training example for a user input. We propose to boost the discriminative ability by transferring a natural language inference (NLI) model. Our extensive experiments on a large-scale multi-domain intent detection task show that our method achieves more stable and accurate in-domain and OOS detection accuracy than RoBERTa-based classifiers and embedding-based nearest neighbor approaches. More notably, the NLI transfer enables our 10-shot model to perform competitively with 50-shot or even full-shot classifiers, while we can keep the inference time constant by leveraging a faster embedding retrieval model., Comment: 19 pages, accepted by EMNLP 2020 main conference as a long paper. Code will be available at https://github.com/salesforce/DNNC-few-shot-intent

We report the synthesis and physical properties studies of quais-1D iron chalcogenide $\rm BaFe_2Se_4$ which shares the $\rm FeSe_4$ tetrahedra building motif commonly seen in the iron chalcogenide superconductors. A high-quality polycrystalline sample was achieved by solid-state reaction method and characterized by X-ray diffraction, electrical resistivity, magnetic susceptibility and neutron diffraction measurements. $\rm BaFe_2Se_4$ is a narrow gap semiconductor that magnetically orders at $\sim$ 310 K. Both neutron powder diffraction results and isothermal M-H loops suggest a canted antiferromagnetic structure where Fe sublattice are antiferromagnetically ordered along the c-axis quasi-1D chain direction, resulting in a net ferromagnetic moment in the perpendicular direction along the a-axis with tilted angle of 18.7$^\circ$ towards the b-axis., Comment: 8 pages, 4 figures