While Large Language Models (LLMs) have found success in real-world applications, their underlying explanatory process is still poorly understood. This paper proposes IBE-Eval, a framework inspired by philosophical accounts on Inference to the Best Explanation (IBE) to advance the interpretation and evaluation of LLMs’ explanations. IBE-Eval estimates the plausibility of natural language explanations through a combination of explicit logical and linguistic features including: consistency, parsimony, coherence, and uncertainty. Extensive experiments are conducted on Causal Question Answering (CQA), where IBE-Eval is tasked to select the most plausible causal explanation amongst competing ones generated by LLMs (i.e., GPT 3.5 and Llama 2). The experiments reveal that IBE-Eval can successfully identify the best explanation with up to 77% accuracy (≈ 27% above random), improving upon a GPT 3.5-as-a-Judge baseline (≈+17%) while being intrinsically more efficient and interpretable. Additional analyses suggest that, despite model-specific variances, LLM-generated explanations tend to conform to IBE criteria and that IBE-Eval is significantly correlated with human judgment, opening up opportunities for future development of automated explanation verification tools.

Disentangled latent spaces usually have better semantic separability and geometrical properties, which leads to better interpretability and more controllable data generation. While this has been well investigated in Computer Vision, in tasks such as image disentanglement, in the NLP domain, sentence disentanglement is still comparatively under-investigated. Most previous work have concentrated on disentangling task-specific generative factors, such as sentiment, within the context of style transfer. In this work, we focus on a more general form of sentence disentanglement, targeting the localised modification and control of more general sentence semantic features. To achieve this, we contribute to a novel notion of sentence semantic disentanglement and introduce a flow-based invertible neural network (INN) mechanism integrated with a transformer-based language Autoencoder (AE) in order to deliver latent spaces with better separability properties. Experimental results demonstrate that the model can conform the distributed latent space into a better semantically disentangled sentence space, leading to improved language interpretability and controlled generation when compared to the recent state-of-the-art language VAE models.

We present BioLunar, developed using the Lunar framework, as a tool for supporting biological analyses, with a particular emphasis on molecular-level evidence enrichment for biomarker discovery in oncology. The platform integrates Large Language Models (LLMs) to facilitate complex scientific reasoning across distributed evidence spaces, enhancing the capability for harmonizing and reasoning over heterogeneous data sources. Demonstrating its utility in cancer research, BioLunar leverages modular design, reusable data access and data analysis components, and a low-code user interface, enabling researchers of all programming levels to construct LLM-enabled scientific workflows. By facilitating automatic scientific discovery and inference from heterogeneous evidence, BioLunar exemplifies the potential of the integration between LLMs, specialised databases and biomedical tools to support expert-level knowledge synthesis and discovery.

The sparsity of labeled data is an obstacle to the development of Relation Extraction (RE) models and the completion of databases in various biomedical areas. While being of high interest in drug-discovery, the literature on natural products, reporting the identification of potential bioactive compounds from organisms, is a concrete example of such an overlooked topic. To mark the start of this new task, we created the first curated evaluation dataset and extracted literature items from the LOTUS database to build training sets. To this end, we developed a new sampler, inspired by diversity metrics in ecology, named Greedy Maximum Entropy sampler (https://github.com/idiap/gme-sampler). The strategic optimization of both balance and diversity of the selected items in the evaluation set is important given the resource-intensive nature of manual curation. After quantifying the noise in the training set, in the form of discrepancies between the text of input abstracts and the expected output labels, we explored different strategies accordingly. Framing the task as an end-to-end Relation Extraction, we evaluated the performance of standard fine-tuning (BioGPT, GPT-2, and Seq2rel) and few-shot learning with open Large Language Models (LLMs) (LLaMA 7B-65B). In addition to their evaluation in few-shot settings, we explore the potential of open LLMs as synthetic data generators and propose a new workflow for this purpose. All evaluated models exhibited substantial improvements when fine-tuned on synthetic abstracts rather than the original noisy data. We provide our best performing (F1-score = 59.0) BioGPT-Large model for end-to-end RE of natural products relationships along with all the training and evaluation datasets. See more details at https://github.com/idiap/abroad-re.

An increasing amount of research in Natural Language Inference (NLI) focuses on the application and evaluation of Large Language Models (LLMs) and their reasoning capabilities. Despite their success, however, LLMs are still prone to factual errors and inconsistencies in their explanations, offering limited control and interpretability for inference in complex domains. In this paper, we focus on ethical NLI, investigating how hybrid neuro-symbolic techniques can enhance the logical validity and alignment of ethical explanations produced by LLMs. Specifically, we present an abductive-deductive framework named Logic-Explainer, which integrates LLMs with an external backward-chaining solver to refine step-wise natural language explanations and jointly verify their correctness, reduce incompleteness and minimise redundancy. An extensive empirical analysis demonstrates that Logic-Explainer can improve explanations generated via in-context learning methods and Chain-of-Thought (CoT) on challenging ethical NLI tasks, while, at the same time, producing formal proofs describing and supporting models’ reasoning. As ethical NLI requires commonsense reasoning to identify underlying moral violations, our results suggest the effectiveness of neuro-symbolic methods for multi-step NLI more broadly, opening new opportunities to enhance the logical consistency, reliability, and alignment of LLMs.

Natural language definitions possess a recursive, self-explanatory semantic structure that can support representation learning methods able to preserve explicit conceptual relations and constraints in the latent space. This paper presents a multi-relational model that explicitly leverages such a structure to derive word embeddings from definitions. By automatically extracting the relations linking defined and defining terms from dictionaries, we demonstrate how the problem of learning word embeddings can be formalised via a translational framework in Hyperbolic space and used as a proxy to capture the global semantic structure of definitions. An extensive empirical analysis demonstrates that the framework can help imposing the desired structural constraints while preserving the semantic mapping required for controllable and interpretable traversal. Moreover, the experiments reveal the superiority of the Hyperbolic word embeddings over the Euclidean counterparts and demonstrate that the multi-relational approach can obtain competitive results when compared to state-of-the-art neural models, with the advantage of being intrinsically more efficient and interpretable

Chain-of-Thought (CoT) prompting empowersthe reasoning abilities of Large Language Models (LLMs), eliciting them to solve complexreasoning tasks in a step-wise manner. However, these capabilities appear only in models with billions of parameters, which represent an entry barrier for many users who are constrained to operate on a smaller model scale, i.e., Small Language Models (SLMs). Although many companies are releasing LLMs of the same family with fewer parameters, these models tend not to preserve all the reasoning capabilities of the original models, including CoT reasoning.In this paper, we propose a method for aligning and transferring reasoning abilities between larger to smaller Language Models. By using an Instruction-tuning-CoT method, that is, an Instruction-tuning designed around CoT-Demonstrations, we enable the SLMs to generate multi-step controlled reasoned answers when they are elicited with the CoT mechanism. Hence, we instruct a smaller Language Model using outputs generated by more robust models belonging to the same family or not, evaluating the impact across different types of models. Results obtained on question-answering and mathematical reasoning benchmarks show that LMs instructed via the Instruction-tuning CoT method produced by LLMs outperform baselines within both in-domain and out-domain scenarios.

The alignment of reasoning abilities between smaller and larger Language Models are largely conducted via supervised fine-tuning using demonstrations generated from robust Large Language Models (LLMs). Although these approaches deliver more performant models, they do not show sufficiently strong generalization ability as the training only relies on the provided demonstrations.In this paper, we propose the Self-refine Instruction-tuning method that elicits Smaller Language Models to self-improve their abilities.Our approach is based on a two-stage process, where reasoning abilities are first transferred between LLMs and Small Language Models (SLMs) via Instruction-tuning on synthetic demonstrations provided by LLMs, and then the instructed models self-improve their abilities through preference optimization strategies.In particular, the second phase operates refinement heuristics based on Direct Preference Optimization, where the SLMs are elicited to deliver a series of reasoning paths by automatically sampling the generated responses and providing rewards using ground truths from the LLMs.Results obtained on commonsense and math reasoning tasks show that this approach consistently outperforms Instruction-tuning in both in-domain and out-domain scenarios, aligning the reasoning abilities of Smaller and Larger language models.

Natural language explanations represent a proxy for evaluating explanation-based and multi-step Natural Language Inference (NLI) models. However, assessing the validity of explanations for NLI is challenging as it typically involves the crowd-sourcing of apposite datasets, a process that is time-consuming and prone to logical errors. To address existing limitations, this paper investigates the verification and refinement of natural language explanations through the integration of Large Language Models (LLMs) and Theorem Provers (TPs). Specifically, we present a neuro-symbolic framework, named Explanation-Refiner, that integrates TPs with LLMs to generate and formalise explanatory sentences and suggest potential inference strategies for NLI. In turn, the TP is employed to provide formal guarantees on the logical validity of the explanations and to generate feedback for subsequent improvements. We demonstrate how Explanation-Refiner can be jointly used to evaluate explanatory reasoning, autoformalisation, and error correction mechanisms of state-of-the-art LLMs as well as to automatically enhance the quality of explanations of variable complexity in different domains.

Autoformalization is the task of automatically translating mathematical content written in natural language to a formal language expression. The growing language interpretation capabilities of Large Language Models (LLMs), including in formal languages, are lowering the barriers for autoformalization. However, LLMs alone are not capable of consistently and reliably delivering autoformalization, in particular as the complexity and specialization of the target domain grows. As the field evolves into the direction of systematically applying autoformalization towards large mathematical libraries, the need to improve syntactic, terminological and semantic control increases. This paper proposes the coordinated use of three mechanisms, most-similar retrieval augmented generation (MS-RAG), denoising steps, and auto-correction with syntax error feedback (Auto-SEF) to improve autoformalization quality. The empirical analysis, across different models, demonstrates that these mechanisms can deliver autoformalizaton results which are syntactically, terminologically and semantically more consistent. These mechanisms can be applied across different LLMs and have shown to deliver improve results across different model types.

Explanation constitutes an archetypal feature of human rationality, underpinning learning and generalisation, and representing one of the media supporting scientific discovery and communication. Due to the importance of explanations in human reasoning, an increasing amount of research in Natural Language Inference (NLI) has started reconsidering the role that explanations play in learning and inference, attempting to build explanation-based NLI models that can effectively encode and use natural language explanations on downstream tasks. Research in explanation-based NLI, however, presents specific challenges and opportunities, as explanatory reasoning reflects aspects of both material and formal inference, making it a particularly rich setting to model and deliver complex reasoning. In this tutorial, we provide a comprehensive introduction to the field of explanation-based NLI, grounding this discussion on the epistemological-linguistic foundations of explanations, systematically describing the main architectural trends and evaluation methodologies that can be used to build systems capable of explanatory reasoning.

Achieving precise semantic control over the latent spaces of Variational AutoEncoders (VAEs) holds significant value for downstream tasks in NLP as the underlying generative mechanisms could be better localised, explained and improved upon. Recent research, however, has struggled to achieve consistent results, primarily due to the inevitable loss of semantic information in the variational bottleneck and limited control over the decoding mechanism. To overcome these challenges, we investigate discrete latent spaces in Vector Quantized Variational AutoEncoder (VQVAE) to improve semantic control and generation in Transformer-based VAEs. In particular, We propose T5VQVAE, a novel model that leverages the controllability of VQVAE to guide the self-attention mechanism in T5, exploiting its full generalization capabilities. Experimental results indicate that T5VQVAE outperforms existing state-of-the-art VAE models, including Optimus, in terms of control and preservation of semantic information across different tasks such as auto-encoding of sentences and mathematical expressions, text transfer, and inference. Moreover, T5VQVAE exhibits improved reasoning capabilities, suggesting potential applications for downstream natural language and symbolic inference tasks.

The injection of syntactic information in Variational AutoEncoders (VAEs) can result in an overall improvement of performances and generalisation. An effective strategy to achieve such a goal is to separate the encoding of distributional semantic features and syntactic structures into heterogeneous latent spaces via multi-task learning or dual encoder architectures. However, existing works employing such techniques are limited to LSTM-based VAEs. This work investigates latent space separation methods for structural syntactic injection in Transformer-based VAE architectures (i.e., Optimus) through the integration of graph-based models. Our empirical evaluation reveals that the proposed end-to-end VAE architecture can improve theoverall organisation of the latent space, alleviating the information loss occurring in standard VAE setups, and resulting in enhanced performances on language modelling and downstream generation tasks.

The language ability of Large Language Models (LLMs) is often unbalanced towards English because of the imbalance in the distribution of the pre-training data. This disparity is demanded in further fine-tuning and affecting the cross-lingual abilities of LLMs. In this paper, we propose to empower Instruction-tuned LLMs (It-LLMs) in languages other than English by building semantic alignment between them. Hence, we propose CrossAlpaca, an It-LLM with cross-lingual Instruction-following and Translation-following demonstrations to improve semantic alignment between languages. We validate our approach on the multilingual Question Answering (QA) benchmarks XQUAD and MLQA and adapted versions of MMLU and BBH.Our models, tested over six different languages, outperform the It-LLMs tuned on monolingual data. The final results show that instruction tuning on non-English data is not enough and that semantic alignment can be further improved by Translation-following demonstrations.

Language models (LMs) can hallucinate when performing complex mathematical reasoning. Physics provides a rich domain for assessing their mathematical capabilities, where physical context requires that any symbolic manipulation satisfies complex semantics (e.g., units, tensorial order). In this work, we systematically remove crucial context from prompts to force instances where model inference may be algebraically coherent, yet unphysical. We assess LM capabilities in this domain using a curated dataset encompassing multiple notations and Physics subdomains. Further, we improve zero-shot scores using synthetic in-context examples, and demonstrate non-linear degradation of derivation quality with perturbation strength via the progressive omission of supporting premises. We find that the models’ mathematical reasoning is not physics-informed in this setting, where physical context is predominantly ignored in favour of reverse-engineering solutions.

Integer Linear Programming (ILP) has been proposed as a formalism for encoding precise structural and semantic constraints for Natural Language Inference (NLI). However, traditional ILP frameworks are non-differentiable, posing critical challenges for the integration of continuous language representations based on deep learning. In this paper, we introduce a novel approach, named Diff-Comb Explainer, a neuro-symbolic architecture for explanation-based NLI based on Differentiable BlackBox Combinatorial Solvers (DBCS). Differently from existing neuro-symbolic solvers, Diff-Comb Explainer does not necessitate a continuous relaxation of the semantic constraints, enabling a direct, more precise, and efficient incorporation of neural representations into the ILP formulation. Our experiments demonstrate that Diff-Comb Explainer achieves superior performance when compared to conventional ILP solvers, neuro-symbolic black-box solvers, and Transformer-based encoders. Moreover, a deeper analysis reveals that Diff-Comb Explainer can significantly improve the precision, consistency, and faithfulness of the constructed explanations, opening new opportunities for research on neuro-symbolic architectures for explainable and transparent NLI in complex domains.

Curriculum Learning (CL) has been emerged as an effective technique for improving the performances and reducing the cost of pre-training Large Language Models (LLMs). The efficacy of CL demonstrated in different scenarios is in the training LLMs by organizing examples from the simplest to the most complex. Although improvements have been shown extensively, this approach was used for pre-training, leaving novel fine-tuning approaches such as instruction-tuning unexplored. In this paper, we propose a novel complexity measure to empower the instruction-tuning method using the CL paradigm. To complement previous works, we propose cognitively motivated measures to determine the complexity of training demonstrations used in the instruction-tuning paradigm. Hence, we experiment with the proposed heuristics first in English and then in other languages. The downstream results show that delivering training examples by complexity ranking is also effective for instruction tuning, as it improves downstream performance while reducing costs. Furthermore, the technique can be easily transferred to languages other than English, e.g., Italian and French, without any adaptation, maintaining functionality and effectiveness.

Rigorous evaluation of the causal effects of semantic features on language model predictions can be hard to achieve for natural language reasoning problems. However, this is such a desirable form of analysis from both an interpretability and model evaluation perspective, that it is valuable to investigate specific patterns of reasoning with enough structure and regularity to identify and quantify systematic reasoning failures in widely-used models. In this vein, we pick a portion of the NLI task for which an explicit causal diagram can be systematically constructed: the case where across two sentences (the premise and hypothesis), two related words/terms occur in a shared context. In this work, we apply causal effect estimation strategies to measure the effect of context interventions (whose effect on the entailment label is mediated by the semantic monotonicity characteristic) and interventions on the inserted word-pair (whose effect on the entailment label is mediated by the relation between these words). Extending related work on causal analysis of NLP models in different settings, we perform an extensive interventional study on the NLI task to investigate robustness to irrelevant changes and sensitivity to impactful changes of Transformers. The results strongly bolster the fact that similar benchmark accuracy scores may be observed for models that exhibit very different behaviour. Moreover, our methodology reinforces previously suspected biases from a causal perspective, including biases in favour of upward-monotone contexts and ignoring the effects of negation markers.

Text embeddings provide a concise representation of the semantics of sentences and larger spans of text, rather than individual words, capturing a wide range of linguistic features. They have found increasing application to a variety of NLP tasks, including machine translation and natural language inference. While most recent breakthroughs in task performance are being achieved by large scale distributional models, there is a growing disconnection between their knowledge representation and traditional semantics, which hinders efforts to capture such knowledge in human interpretable form or explain model inference behaviour. In this tutorial, we examine from basics to the cutting edge research on the analysis and control of text representations, aiming to shorten the gap between deep latent semantics and formal symbolics. This includes the considerations on knowledge formalisation, the linguistic information that can be extracted and measured from distributional models, and intervention techniques that enable explainable reasoning and controllable text generation, covering methods from pooling to LLM-based.

This paper investigates the possibility of approximating multiple mathematical operations in latent space for expression derivation. To this end, we introduce different multi-operational representation paradigms, modelling mathematical operations as explicit geometric transformations. By leveraging a symbolic engine, we construct a large-scale dataset comprising 1.7M derivation steps stemming from 61K premises and 6 operators, analysing the properties of each paradigm when instantiated with state-of-the-art neural encoders.Specifically, we investigate how different encoding mechanisms can approximate expression manipulation in latent space, exploring the trade-off between learning different operators and specialising within single operations, as well as the ability to support multi-step derivations and out-of-distribution generalisation. Our empirical analysis reveals that the multi-operational paradigm is crucial for disentangling different operators, while discriminating the conclusions for a single operation is achievable in the original expression encoder. Moreover, we show that architectural choices can heavily affect the training dynamics, structural organisation, and generalisation of the latent space, resulting in significant variations across paradigms and classes of encoders.

This paper proposes a methodology for generating and perturbing detailed derivations of equations at scale, aided by a symbolic engine, to evaluate the generalisability of Transformers to out-of-distribution mathematical reasoning problems. Instantiating the framework in the context of sequence classification tasks, we compare the capabilities of GPT-4, GPT-3.5, and a canon of fine-tuned BERT models, exploring the relationship between specific operators and generalisation failure via the perturbation of reasoning aspects such as symmetry and variable surface forms. Surprisingly, our empirical evaluation reveals that the average in-distribution performance of fine-tuned models surpasses GPT-3.5, and rivals GPT-4. However, perturbations to input reasoning can reduce their performance by up to 80 F1 points. Overall, the results suggest that the in-distribution performance of smaller open-source models may potentially rival GPT by incorporating appropriately structured derivation dependencies during training, and highlight a shared weakness between BERT and GPT involving a relative inability to decode indirect references to mathematical entities. We release the full codebase, constructed datasets, and fine-tuned models to encourage future progress in the field.

Large Language Models (LLMs) are at the forefront of NLP achievements but fall short in dealing with shortcut learning, factual inconsistency, and vulnerability to adversarial inputs. These shortcomings are especially critical in medical contexts, where they can misrepresent actual model capabilities. Addressing this, we present SemEval-2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials. Our contributions include the refined NLI4CT-P dataset (i.e. Natural Language Inference for Clinical Trials - Perturbed), designed to challenge LLMs with interventional and causal reasoning tasks, along with a comprehensive evaluation of methods and results for participant submissions. A total of 106 participants registered for the task contributing to over 1200 individual submissions and 25 system overview papers. This initiative aims to advance the robustness and applicability of NLI models in healthcare, ensuring safer and more dependable AI assistance in clinical decision-making. We anticipate that the dataset, models, and outcomes of this task can support future research in the field of biomedical NLI. The dataset, competition leaderboard, and website are publicly available.

Specialized transformers-based models (such as BioBERT and BioMegatron) are adapted for the biomedical domain based on publicly available biomedical corpora. As such, they have the potential to encode large-scale biological knowledge. We investigate the encoding and representation of biological knowledge in these models, and its potential utility to support inference in cancer precision medicine—namely, the interpretation of the clinical significance of genomic alterations. We compare the performance of different transformer baselines; we use probing to determine the consistency of encodings for distinct entities; and we use clustering methods to compare and contrast the internal properties of the embeddings for genes, variants, drugs, and diseases. We show that these models do indeed encode biological knowledge, although some of this is lost in fine-tuning for specific tasks. Finally, we analyze how the models behave with regard to biases and imbalances in the dataset.

How can we interpret and retrieve medical evidence to support clinical decisions? Clinical trial reports (CTR) amassed over the years contain indispensable information for the development of personalized medicine. However, it is practically infeasible to manually inspect over 400,000+ clinical trial reports in order to find the best evidence for experimental treatments. Natural Language Inference (NLI) offers a potential solution to this problem, by allowing the scalable computation of textual entailment. However, existing NLI models perform poorly on biomedical corpora, and previously published datasets fail to capture the full complexity of inference over CTRs. In this work, we present a novel resource to advance research on NLI for reasoning on CTRs. The resource includes two main tasks. Firstly, to determine the inference relation between a natural language statement, and a CTR. Secondly, to retrieve supporting facts to justify the predicted relation. We provide NLI4CT, a corpus of 2400 statements and CTRs, annotated for these tasks. Baselines on this corpus expose the limitations of existing NLI approaches, with 6 state-of-the-art NLI models achieving a maximum F1 score of 0.627. To the best of our knowledge, we are the first to design a task that covers the interpretation of full CTRs. To encourage further work on this challenging dataset, we make the corpus, competition leaderboard, and website, available on CodaLab, and code to replicate the baseline experiments on GitHub.

Disentangling the encodings of neural models is a fundamental aspect for improving interpretability, semantic control and downstream task performance in Natural Language Processing. Currently, most disentanglement methods are unsupervised or rely on synthetic datasets with known generative factors. We argue that recurrent syntactic and semantic regularities in textual data can be used to provide the models with both structural biases and generative factors. We leverage the semantic structures present in a representative and semantically dense category of sentence types, definitional sentences, for training a Variational Autoencoder to learn disentangled representations. Our experimental results show that the proposed model outperforms unsupervised baselines on several qualitative and quantitative benchmarks for disentanglement, and it also improves the results in the downstream task of definition modeling.

Probing strategies have been shown to detectthe presence of various linguistic features inlarge language models; in particular, seman-tic features intermediate to the “natural logic”fragment of the Natural Language Inferencetask (NLI). In the case of natural logic, the rela-tion between the intermediate features and theentailment label is explicitly known: as such,this provides a ripe setting for interventionalstudies on the NLI models’ representations, al-lowing for stronger causal conjectures and adeeper critical analysis of interventional prob-ing methods. In this work, we carry out newand existing representation-level interventionsto investigate the effect of these semantic fea-tures on NLI classification: we perform am-nesic probing (which removes features as di-rected by learned linear probes) and introducethe mnestic probing variation (which forgetsall dimensions except the probe-selected ones).Furthermore, we delve into the limitations ofthese methods and outline some pitfalls havebeen obscuring the effectivity of interventionalprobing studies.

This paper describes the results of SemEval 2023 task 7 – Multi-Evidence Natural Language Inference for Clinical Trial Data (NLI4CT) – consisting of 2 tasks, a Natural Language Inference (NLI) task, and an evidence selection task on clinical trial data. The proposed challenges require multi-hop biomedical and numerical reasoning, which are of significant importance to the development of systems capable of large-scale interpretation and retrieval of medical evidence, to provide personalized evidence-based care. Task 1, the entailment task, received 643 submissions from 40 participants, and Task 2, the evidence selection task, received 364 submissions from 23 participants. The tasks are challenging, with the majority of submitted systems failing to significantly outperform the majority class baseline on the entailment task, and we observe significantly better performance on the evidence selection task than on the entailment task. Increasing the number of model parameters leads to a direct increase in performance, far more significant than the effect of biomedical pre-training. Future works could explore the limitations of large models for generalization and numerical inference, and investigate methods to augment clinical datasets to allow for more rigorous testing and to facilitate fine-tuning. We envisage that the dataset, models, and results of this task will be useful to the biomedical NLI and evidence retrieval communities. The dataset, competition leaderboard, and website are publicly available.

Automating discovery in mathematics and science will require sophisticated methods of information extraction and abstract reasoning, including models that can convincingly process relationships between mathematical elements and natural language, to produce problem solutions of real-world value. We analyze mathematical language processing methods across five strategic sub-areas (identifier-definition extraction, formula retrieval, natural language premise selection, math word problem solving, and informal theorem proving) from recent years, highlighting prevailing methodologies, existing limitations, overarching trends, and promising avenues for future research.

In the interest of interpreting neural NLI models and their reasoning strategies, we carry out a systematic probing study which investigates whether these modelscapture the crucial semantic features central to natural logic: monotonicity and concept inclusion. Correctly identifying valid inferences in downward-monotone contexts is a known stumbling block for NLI performance,subsuming linguistic phenomena such as negation scope and generalized quantifiers. To understand this difficulty, we emphasize monotonicity as a property of a context and examine the extent to which models capture relevant monotonicity information in the vector representations which are intermediate to their decision making process. Drawing on the recent advancement of the probing paradigm,we compare the presence of monotonicity features across various models. We find that monotonicity information is notably weak in the representations of popularNLI models which achieve high scores on benchmarks, and observe that previous improvements to these models based on fine-tuning strategies have introduced stronger monotonicity features together with their improved performance on challenge sets.

We present a discourse-aware text simplification (TS) approach that recursively splits and rephrases complex English sentences into a semantic hierarchy of simplified sentences. Using a set of linguistically principled transformation patterns, sentences are converted into a hierarchical representation in the form of core sentences and accompanying contexts that are linked via rhetorical relations. As opposed to previously proposed sentence splitting approaches, which commonly do not take into account discourse-level aspects, our TS approach preserves the semantic relationship of the decomposed constituents in the output. A comparative analysis with the annotations contained in RST-DT shows that we capture the contextual hierarchy between the split sentences with a precision of 89% and reach an average precision of 69% for the classification of the rhetorical relations that hold between them. Moreover, an integration into state-of-the-art Open Information Extraction (IE) systems reveals that when applying our TS approach as a pre-processing step, the generated relational tuples are enriched with additional meta information, resulting in a novel lightweight semantic representation for the task of Open IE.

Most of the contemporary approaches for multi-hop Natural Language Inference (NLI) construct explanations considering each test case in isolation. However, this paradigm is known to suffer from semantic drift, a phenomenon that causes the construction of spurious explanations leading to wrong conclusions. In contrast, this paper proposes an abductive framework for multi-hop NLI exploring the retrieve-reuse-refine paradigm in Case-Based Reasoning (CBR). Specifically, we present Case-Based Abductive Natural Language Inference (CB-ANLI), a model that addresses unseen inference problems by analogical transfer of prior explanations from similar examples. We empirically evaluate the abductive framework on commonsense and scientific question answering tasks, demonstrating that CB-ANLI can be effectively integrated with sparse and dense pre-trained encoders to improve multi-hop inference, or adopted as an evidence retriever for Transformers. Moreover, an empirical analysis of semantic drift reveals that the CBR paradigm boosts the quality of the most challenging explanations, a feature that has a direct impact on robustness and accuracy in downstream inference tasks.

The application of Natural Language Inference (NLI) methods over large textual corpora can facilitate scientific discovery, reducing the gap between current research and the available large-scale scientific knowledge. However, contemporary NLI models are still limited in interpreting mathematical knowledge written in Natural Language, even though mathematics is an integral part of scientific argumentation for many disciplines. One of the fundamental requirements towards mathematical language understanding, is the creation of models able to meaningfully represent variables. This problem is particularly challenging since the meaning of a variable should be assigned exclusively from its defining type, i.e., the representation of a variable should come from its context. Recent research has formalised the variable typing task, a benchmark for the understanding of abstract mathematical types and variables in a sentence. In this work, we propose VarSlot, a Variable Slot-based approach, which not only delivers state-of-the-art results in the task of variable typing, but is also able to create context-based representations for variables.

Metamorphic testing has recently been used to check the safety of neural NLP models. Its main advantage is that it does not rely on a ground truth to generate test cases. However, existing studies are mostly concerned with robustness-like metamorphic relations, limiting the scope of linguistic properties they can test. We propose three new classes of metamorphic relations, which address the properties of systematicity, compositionality and transitivity. Unlike robustness, our relations are defined over multiple source inputs, thus increasing the number of test cases that we can produce by a polynomial factor. With them, we test the internal consistency of state-of-the-art NLP models, and show that they do not always behave according to their expected linguistic properties. Lastly, we introduce a novel graphical notation that efficiently summarises the inner structure of metamorphic relations.

In order for language models to aid physics research, they must first encode representations of mathematical and natural language discourse which lead to coherent explanations, with correct ordering and relevance of statements. We present a collection of datasets developed to evaluate the performance of language models in this regard, which measure capabilities with respect to sentence ordering, position, section prediction, and discourse coherence. Analysis of the data reveals the classes of arguments and sub-disciplines which are most common in physics discourse, as well as the sentence-level frequency of equations and expressions. We present baselines that demonstrate how contemporary language models are challenged by coherence related tasks in physics, even when trained on mathematical natural language objectives.

We present a case study on the application of text classification and legal judgment prediction for flight compensation. We combine transformer-based classification models to classify responses from airlines and incorporate text data with other data types to predict a legal claim being successful. Our experimental evaluations show that our models achieve consistent and significant improvements over baselines and even outperformed human prediction when predicting a claim being successful. These models were integrated into an existing claim management system, providing substantial productivity gains for handling the case lifecycle, currently supporting several thousands of monthly processes.

Reinforcement Learning has shown success in a number of complex virtual environments. However, many challenges still exist towards solving problems with natural language as a core component. Interactive Fiction Games (or Text Games) are one such problem type that offer a set of safe, partially observable environments where natural language is required as part of the Reinforcement Learning solution. Therefore, this survey’s aim is to assist in the development of new Text Game problem settings and solutions for Reinforcement Learning informed by natural language. Specifically, this survey: 1) introduces the challenges in Text Game Reinforcement Learning problems, 2) outlines the generation tools for rendering Text Games and the subsequent environments generated, and 3) compares the agent architectures currently applied to provide a systematic review of benchmark methodologies and opportunities for future researchers.

This paper presents Diff-Explainer, the first hybrid framework for explainable multi-hop inference that integrates explicit constraints with neural architectures through differentiable convex optimization. Specifically, Diff- Explainer allows for the fine-tuning of neural representations within a constrained optimization framework to answer and explain multi-hop questions in natural language. To demonstrate the efficacy of the hybrid framework, we combine existing ILP-based solvers for multi-hop Question Answering (QA) with Transformer-based representations. An extensive empirical evaluation on scientific and commonsense QA tasks demonstrates that the integration of explicit constraints in a end-to-end differentiable framework can significantly improve the performance of non- differentiable ILP solvers (8.91%–13.3%). Moreover, additional analysis reveals that Diff-Explainer is able to achieve strong performance when compared to standalone Transformers and previous multi-hop approaches while still providing structured explanations in support of its predictions.

The Shared Task on Natural Language Premise Selection (NLPS) asks participants to retrieve the set of premises that are most likely to be useful for proving a given mathematical statement from a supporting knowledge base. While previous editions of the TextGraphs shared tasks series targeted multi-hop inference for explanation regeneration in the context of science questions (Thayaparan et al., 2021; Jansen and Ustalov, 2020, 2019), NLPS aims to assess the ability of state-of-the-art approaches to operate on a mixture of natural and mathematical language and model complex multi-hop reasoning dependencies between statements. To this end, this edition of the shared task makes use of a large set of approximately 21k mathematical statements extracted from the PS-ProofWiki dataset (Ferreira and Freitas, 2020a). In this summary paper, we present the results of the 1st edition of the NLPS task, providing a description of the evaluation data, and the participating systems. Additionally, we perform a detailed analysis of the results, evaluating various aspects involved in mathematical language processing and multi-hop inference. The best-performing system achieved a MAP of 15.39, improving the performance of a TF-IDF baseline by approximately 3.0 MAP.

Probing (or diagnostic classification) has become a popular strategy for investigating whether a given set of intermediate features is present in the representations of neural models. Naive probing studies may have misleading results, but various recent works have suggested more reliable methodologies that compensate for the possible pitfalls of probing. However, these best practices are numerous and fast-evolving. To simplify the process of running a set of probing experiments in line with suggested methodologies, we introduce Probe-Ably: an extendable probing framework which supports and automates the application of probing methods to the user’s inputs.

This paper presents a novel framework for reconstructing multi-hop explanations in science Question Answering (QA). While existing approaches for multi-hop reasoning build explanations considering each question in isolation, we propose a method to leverage explanatory patterns emerging in a corpus of scientific explanations. Specifically, the framework ranks a set of atomic facts by integrating lexical relevance with the notion of unification power, estimated analysing explanations for similar questions in the corpus. An extensive evaluation is performed on the Worldtree corpus, integrating k-NN clustering and Information Retrieval (IR) techniques. We present the following conclusions: (1) The proposed method achieves results competitive with Transformers, yet being orders of magnitude faster, a feature that makes it scalable to large explanatory corpora (2) The unification-based mechanism has a key role in reducing semantic drift, contributing to the reconstruction of many hops explanations (6 or more facts) and the ranking of complex inference facts (+12.0 Mean Average Precision) (3) Crucially, the constructed explanations can support downstream QA models, improving the accuracy of BERT by up to 10% overall.

Mathematical statements written in natural language are usually composed of two different modalities: mathematical elements and natural language. These two modalities have several distinct linguistic and semantic properties. State-of-the-art representation techniques have demonstrated an inability in capturing such an entangled style of discourse. In this work, we propose STAR, a model that uses cross-modal attention to learn how to represent mathematical text for the task of Natural Language Premise Selection. This task uses conjectures written in both natural and mathematical language to recommend premises that most likely will be relevant to prove a particular statement. We found that STAR not only outperforms baselines that do not distinguish between natural language and mathematical elements, but it also achieves better performance than state-of-the-art models.

SemEval is the primary venue in the NLP community for the proposal of new challenges and for the systematic empirical evaluation of NLP systems. This paper provides a systematic quantitative analysis of SemEval aiming to evidence the patterns of the contributions behind SemEval. By understanding the distribution of task types, metrics, architectures, participation and citations over time we aim to answer the question on what is being evaluated by SemEval.

The ability of learning disentangled representations represents a major step for interpretable NLP systems as it allows latent linguistic features to be controlled. Most approaches to disentanglement rely on continuous variables, both for images and text. We argue that despite being suitable for image datasets, continuous variables may not be ideal to model features of textual data, due to the fact that most generative factors in text are discrete. We propose a Variational Autoencoder based method which models language features as discrete variables and encourages independence between variables for learning disentangled representations. The proposed model outperforms continuous and discrete baselines on several qualitative and quantitative benchmarks for disentanglement as well as on a text style transfer downstream application.

This paper explores the topic of transportability, as a sub-area of generalisability. By proposing the utilisation of metrics based on well-established statistics, we are able to estimate the change in performance of NLP models in new contexts. Defining a new measure for transportability may allow for better estimation of NLP system performance in new domains, and is crucial when assessing the performance of NLP systems in new tasks and domains. Through several instances of increasing complexity, we demonstrate how lightweight domain similarity measures can be used as estimators for the transportability in NLP applications. The proposed transportability measures are evaluated in the context of Named Entity Recognition and Natural Language Inference tasks.

This paper describes N-XKT (Neural encoding based on eXplanatory Knowledge Transfer), a novel method for the automatic transfer of explanatory knowledge through neural encoding mechanisms. We demonstrate that N-XKT is able to improve accuracy and generalization on science Question Answering (QA). Specifically, by leveraging facts from background explanatory knowledge corpora, the N-XKT model shows a clear improvement on zero-shot QA. Furthermore, we show that N-XKT can be fine-tuned on a target QA dataset, enabling faster convergence and more accurate results. A systematic analysis is conducted to quantitatively analyze the performance of the N-XKT model and the impact of different categories of knowledge on the zero-shot generalization task.

An emerging line of research in Explainable NLP is the creation of datasets enriched with human-annotated explanations and rationales, used to build and evaluate models with step-wise inference and explanation generation capabilities. While human-annotated explanations are used as ground-truth for the inference, there is a lack of systematic assessment of their consistency and rigour. In an attempt to provide a critical quality assessment of Explanation Gold Standards (XGSs) for NLI, we propose a systematic annotation methodology, named Explanation Entailment Verification (EEV), to quantify the logical validity of human-annotated explanations. The application of EEV on three mainstream datasets reveals the surprising conclusion that a majority of the explanations, while appearing coherent on the surface, represent logically invalid arguments, ranging from being incomplete to containing clearly identifiable logical errors. This conclusion confirms that the inferential properties of explanations are still poorly formalised and understood, and that additional work on this line of research is necessary to improve the way Explanation Gold Standards are constructed.

Natural language contexts display logical regularities with respect to substitutions of related concepts: these are captured in a functional order-theoretic property called monotonicity. For a certain class of NLI problems where the resulting entailment label depends only on the context monotonicity and the relation between the substituted concepts, we build on previous techniques that aim to improve the performance of NLI models for these problems, as consistent performance across both upward and downward monotone contexts still seems difficult to attain even for state of the art models. To this end, we reframe the problem of context monotonicity classification to make it compatible with transformer-based pre-trained NLI models and add this task to the training pipeline. Furthermore, we introduce a sound and complete simplified monotonicity logic formalism which describes our treatment of contexts as abstract units. Using the notions in our formalism, we adapt targeted challenge sets to investigate whether an intermediate context monotonicity classification task can aid NLI models’ performance on examples exhibiting monotonicity reasoning.

The discovery of supporting evidence for addressing complex mathematical problems is a semantically challenging task, which is still unexplored in the field of natural language processing for mathematical text. The natural language premise selection task consists in using conjectures written in both natural language and mathematical formulae to recommend premises that most likely will be useful to prove a particular statement. We propose an approach to solve this task as a link prediction problem, using Deep Convolutional Graph Neural Networks. This paper also analyses how different baselines perform in this task and shows that a graph structure can provide higher F1-score, especially when considering multi-hop premise selection.

Mathematical text is written using a combination of words and mathematical expressions. This combination, along with a specific way of structuring sentences makes it challenging for state-of-art NLP tools to understand and reason on top of mathematical discourse. In this work, we propose a new NLP task, the natural premise selection, which is used to retrieve supporting definitions and supporting propositions that are useful for generating an informal mathematical proof for a particular statement. We also make available a dataset, NL-PS, which can be used to evaluate different approaches for the natural premise selection task. Using different baselines, we demonstrate the underlying interpretation challenges associated with the task.

Machine Reading Comprehension (MRC) is the task of answering a question over a paragraph of text. While neural MRC systems gain popularity and achieve noticeable performance, issues are being raised with the methodology used to establish their performance, particularly concerning the data design of gold standards that are used to evaluate them. There is but a limited understanding of the challenges present in this data, which makes it hard to draw comparisons and formulate reliable hypotheses. As a first step towards alleviating the problem, this paper proposes a unifying framework to systematically investigate the present linguistic features, required reasoning and background knowledge and factual correctness on one hand, and the presence of lexical cues as a lower bound for the requirement of understanding on the other hand. We propose a qualitative annotation schema for the first and a set of approximative metrics for the latter. In a first application of the framework, we analyse modern MRC gold standards and present our findings: the absence of features that contribute towards lexical ambiguity, the varying factual correctness of the expected answers and the presence of lexical cues, all of which potentially lower the reading comprehension complexity and quality of the evaluation data.

Identifying what is at the center of the meaning of a word and what discriminates it from other words is a fundamental natural language inference task. This paper describes an explicit word vector representation model (WVM) to support the identification of discriminative attributes. A core contribution of the paper is a quantitative and qualitative comparative analysis of different types of data sources and Knowledge Bases in the construction of explainable and explicit WVMs: (i) knowledge graphs built from dictionary definitions, (ii) entity-attribute-relationships graphs derived from images and (iii) commonsense knowledge graphs. Using a detailed quantitative and qualitative analysis, we demonstrate that these data sources have complementary semantic aspects, supporting the creation of explicit semantic vector spaces. The explicit vector spaces are evaluated using the task of discriminative attribute identification, showing comparable performance to the state-of-the-art systems in the task (F1-score = 0.69), while delivering full model transparency and explainability.

Recent advances in reading comprehension have resulted in models that surpass human performance when the answer is contained in a single, continuous passage of text. However, complex Question Answering (QA) typically requires multi-hop reasoning - i.e. the integration of supporting facts from different sources, to infer the correct answer. This paper proposes Document Graph Network (DGN), a message passing architecture for the identification of supporting facts over a graph-structured representation of text. The evaluation on HotpotQA shows that DGN obtains competitive results when compared to a reading comprehension baseline operating on raw text, confirming the relevance of structured representations for supporting multi-hop reasoning.

This paper describes DBee, a database to support the construction of data-intensive AI applications. DBee provides a unique data model which operates jointly over large-scale knowledge graphs (KGs) and embedding vector spaces (VSs). This model supports queries which exploit the semantic properties of both types of representations (KGs and VSs). Additionally, DBee aims to facilitate the construction of KGs and VSs, by providing a library of generators, which can be used to create, integrate and transform data into KGs and VSs.

We present an approach for recursively splitting and rephrasing complex English sentences into a novel semantic hierarchy of simplified sentences, with each of them presenting a more regular structure that may facilitate a wide variety of artificial intelligence tasks, such as machine translation (MT) or information extraction (IE). Using a set of hand-crafted transformation rules, input sentences are recursively transformed into a two-layered hierarchical representation in the form of core sentences and accompanying contexts that are linked via rhetorical relations. In this way, the semantic relationship of the decomposed constituents is preserved in the output, maintaining its interpretability for downstream applications. Both a thorough manual analysis and automatic evaluation across three datasets from two different domains demonstrate that the proposed syntactic simplification approach outperforms the state of the art in structural text simplification. Moreover, an extrinsic evaluation shows that when applying our framework as a preprocessing step the performance of state-of-the-art Open IE systems can be improved by up to 346% in precision and 52% in recall. To enable reproducible research, all code is provided online.

We compiled a new sentence splitting corpus that is composed of 203K pairs of aligned complex source and simplified target sentences. Contrary to previously proposed text simplification corpora, which contain only a small number of split examples, we present a dataset where each input sentence is broken down into a set of minimal propositions, i.e. a sequence of sound, self-contained utterances with each of them presenting a minimal semantic unit that cannot be further decomposed into meaningful propositions. This corpus is useful for developing sentence splitting approaches that learn how to transform sentences with a complex linguistic structure into a fine-grained representation of short sentences that present a simple and more regular structure which is easier to process for downstream applications and thus facilitates and improves their performance.

We introduce DisSim, a discourse-aware sentence splitting framework for English and German whose goal is to transform syntactically complex sentences into an intermediate representation that presents a simple and more regular structure which is easier to process for downstream semantic applications. For this purpose, we turn input sentences into a two-layered semantic hierarchy in the form of core facts and accompanying contexts, while identifying the rhetorical relations that hold between them. In that way, we preserve the coherence structure of the input and, hence, its interpretability for downstream tasks.

We present an Open Information Extraction (IE) approach that uses a two-layered transformation stage consisting of a clausal disembedding layer and a phrasal disembedding layer, together with rhetorical relation identification. In that way, we convert sentences that present a complex linguistic structure into simplified, syntactically sound sentences, from which we can extract propositions that are represented in a two-layered hierarchy in the form of core relational tuples and accompanying contextual information which are semantically linked via rhetorical relations. In a comparative evaluation, we demonstrate that our reference implementation Graphene outperforms state-of-the-art Open IE systems in the construction of correct n-ary predicate-argument structures. Moreover, we show that existing Open IE approaches can benefit from the transformation process of our framework.

We provide a detailed overview of the various approaches that were proposed to date to solve the task of Open Information Extraction. We present the major challenges that such systems face, show the evolution of the suggested approaches over time and depict the specific issues they address. In addition, we provide a critique of the commonly applied evaluation procedures for assessing the performance of Open IE systems and highlight some directions for future work.

We introduce Graphene, an Open IE system whose goal is to generate accurate, meaningful and complete propositions that may facilitate a variety of downstream semantic applications. For this purpose, we transform syntactically complex input sentences into clean, compact structures in the form of core facts and accompanying contexts, while identifying the rhetorical relations that hold between them in order to maintain their semantic relationship. In that way, we preserve the context of the relational tuples extracted from a source sentence, generating a novel lightweight semantic representation for Open IE that enhances the expressiveness of the extracted propositions.

This paper discusses the “Fine-Grained Sentiment Analysis on Financial Microblogs and News” task as part of SemEval-2017, specifically under the “Detecting sentiment, humour, and truth” theme. This task contains two tracks, where the first one concerns Microblog messages and the second one covers News Statements and Headlines. The main goal behind both tracks was to predict the sentiment score for each of the mentioned companies/stocks. The sentiment scores for each text instance adopted floating point values in the range of -1 (very negative/bearish) to 1 (very positive/bullish), with 0 designating neutral sentiment. This task attracted a total of 32 participants, with 25 participating in Track 1 and 29 in Track 2.

This task proposes a challenge to support the interaction between users and applications, micro-services and software APIs using natural language. The task aims for supporting the evaluation and evolution of the discussions surrounding the natural language processing approaches within the context of end-user natural language programming, under scenarios of high semantic heterogeneity/gap.

We present a text simplification approach that is directed at improving the performance of state-of-the-art Open Relation Extraction (RE) systems. As syntactically complex sentences often pose a challenge for current Open RE approaches, we have developed a simplification framework that performs a pre-processing step by taking a single sentence as input and using a set of syntactic-based transformation rules to create a textual input that is easier to process for subsequently applied Open RE systems.

Lately, with the success of Deep Learning techniques in some computational linguistics tasks, many researchers want to explore new models for their linguistics applications. These models tend to be very different from what standard Neural Networks look like, limiting the possibility to use standard Neural Networks frameworks. This work presents NNBlocks, a new framework written in Python to build and train Neural Networks that are not constrained by a specific kind of architecture, making it possible to use it in computational linguistics.

The identification of semantic relations between terms within texts is a fundamental task in Natural Language Processing which can support applications requiring a lightweight semantic interpretation model. Currently, semantic relation classification concentrates on relations which are evaluated over open-domain data. This work provides a critique on the set of abstract relations used for semantic relation classification with regard to their ability to express relationships between terms which are found in a domain-specific corpora. Based on this analysis, this work proposes an alternative semantic relation model based on reusing and extending the set of abstract relations present in the DOLCE ontology. The resulting set of relations is well grounded, allows to capture a wide range of relations and could thus be used as a foundation for automatic classification of semantic relations.

Understanding the semantic relationships between terms is a fundamental task in natural language processing applications. While structured resources that can express those relationships in a formal way, such as ontologies, are still scarce, a large number of linguistic resources gathering dictionary definitions is becoming available, but understanding the semantic structure of natural language definitions is fundamental to make them useful in semantic interpretation tasks. Based on an analysis of a subset of WordNet’s glosses, we propose a set of semantic roles that compose the semantic structure of a dictionary definition, and show how they are related to the definition’s syntactic configuration, identifying patterns that can be used in the development of information extraction frameworks and semantic models.