This paper presents Adept-SQL, a domain-adapted Text2SQL system that addresses critical deployment challenges in professional fields. While modern LLM-based solutions excel on academic benchmarks, we identify three persistent limitations in industrial application: domain-specific knowledge barriers, the schemas complexity in the real world, and the prohibitive computational costs of large LLMs. Our framework introduces two key innovations: a three-stage grounding mechanism combining dynamic terminology expansion, focused schema alignment, and historical query retrieval; coupled with a hybrid prompting architecture that decomposes SQL generation into schema-aware hinting, term disambiguation, and few-shot example incorporation phases. This approach enables efficient execution using smaller open-source LLMs while maintaining semantic precision. Deployed in petroleum engineering domains, our system achieves 97% execution accuracy on real-world databases, demonstrating 49% absolute improvement over SOTA baselines. We release implementation code to advance research in professional Text2SQL systems.

Packing for Supervised Fine-Tuning (SFT) in autoregressive models involves concatenating data points of varying lengths until reaching the designed maximum length to facilitate GPU processing. However, randomly concatenating data points can lead to cross-contamination of sequences due to the significant difference in their subject matter. The mainstream approaches in SFT ensure that each token in the attention calculation phase only focuses on tokens within its own short sequence, without providing additional learning signals for the preceding context. To address these challenges, we introduce Threshold Filtering Packing (TFP), a method that selects samples with related context while maintaining sufficient diversity within the same pack. Our experiments show that TFP offers a simple-to-implement and scalable approach that significantly enhances SFT performance, with observed improvements of up to 7% on GSM8K, 4% on HumanEval. Furthermore, results from bias benchmark datasets highlight TFP’s promising performance in improving fairness while also boosting prediction accuracy by 15%.

Countless decisions shape our lives, and it is crucial to understand the how and why behind them. In this paper, we introduce a new LLM decision-making framework called STRUX, which enhances LLM decision-making by providing structured explanations. These include favorable and adverse facts related to the decision, along with their respective strengths. STRUX begins by distilling lengthy information into a concise table of key facts. It then employs a series of self-reflection steps to determine which of these facts are pivotal, categorizing them as either favorable or adverse in relation to a specific decision. Lastly, we fine-tune an LLM to identify and prioritize these key facts to optimize decision-making. STRUX has been evaluated on the challenging task of forecasting stock investment decisions based on earnings call transcripts and demonstrated superior performance against strong baselines. It enhances decision transparency by allowing users to understand the impact of different factors, representing a meaningful step towards practical decision-making with LLMs.

Representing query translation into relevant entities is a critical component of an infor- mation retrieval system. This paper proposes an unsupervised framework, ESAQueryRank, designed to process natural language queries by mapping n-gram phrases to Wikipedia ti- tles and ranking potential entity and phrase combinations using Explicit Semantic Analy- sis. Unlike previous approaches, this frame- work does not rely on query expansion, syn- tactic parsing, or manual annotation. Instead, it leverages Wikipedia metadata—such as ti- tles, redirects, disambiguation pages to dis- ambiguate entities and identify the most rel- evant ones based on cosine similarity in the ESA space. ESAQueryRank is evaluated using a random set of TREC questions and compared against a keyword-based approach and a context-based question translation model (CBQT). In all comparisons of full category types, ESAQueryRank consistently shows bet- ter results against both methods. Notably, the framework excels with more complex queries, achieving improvements in Mean Reciprocal Rank (MRR) of up to 480% for intricate queries like those beginning with “Why,” even without explicitly incorporating the question type. These results demonstrate that ESA- QueryRank is an effective, transparent, and domain-independent framework for building natural language interfaces.

Clinical natural language processing faces challenges like complex medical terminology and clinical contexts. Recently, large language models (LLMs) have shown promise in this domain. Yet, their direct deployment can lead to privacy issues and are constrained by resources. To address this challenge, we delve into synthetic clinical text generation with LLMs for clinical NLP tasks. We propose an innovative, resource-efficient approach, ClinGen, which infuses knowledge into the process. Our model involves clinical knowledge extraction and context-informed LLM prompting. Both clinical topics and writing styles are drawn from external domain-specific knowledge graphs and LLMs to guide data generation. Our extensive empirical study across 8 clinical NLP tasks and 18 datasets reveals that ClinGen consistently enhances performance across various tasks by 7.7%-8.7% on average, effectively aligning the distribution of real datasets and enriching the diversity of generated training instances.

Tweets are short messages that often include specialized language such as hashtags and emojis. In this paper, we present a simple strategy to process emojis: replace them with their natural language description and use pretrained word embeddings as normally done with standard words. We show that this strategy is more effective than using pretrained emoji embeddings for tweet classification. Specifically, we obtain new state-of-the-art results in irony detection and sentiment analysis despite our neural network is simpler than previous proposals.