Yun Chen

Nan Huo, Xiaohan Xu, Jinyang Li et al.

Large language models (LLMs) have demonstrated remarkable performance on single-turn text-to-SQL tasks, but real-world database applications predominantly require multi-turn interactions to handle ambiguous queries, execution errors, and evolving user requirements. Existing multi-turn benchmarks fall short by treating conversation histories as static context or limiting evaluation to read-only operations, failing to reflect production-grade database assistant challenges. We introduce BIRD-INTERACT, a benchmark that restores this realism through: (1) a comprehensive interaction environment coupling each database with a hierarchical knowledge base, metadata files, and a function-driven user simulator, enabling models to solicit clarifications, retrieve knowledge, and recover from errors without human supervision; (2) two evaluation settings consisting of a pre-defined conversational protocol (c-Interact) and an open-ended agentic setting (a-Interact) where models autonomously decide when to query the user simulator or explore the environment; (3) a challenging task suite covering the full CRUD spectrum for business-intelligence and operational use cases, guarded by executable test cases. Each task features ambiguous and follow-up sub-tasks requiring dynamic interaction. The suite comprises BIRD-INTERACT-FULL (600 tasks, up to 11,796 interactions) for comprehensive performance assessment, and BIRD-INTERACT-LITE (300 tasks with simplified databases) for detailed behavioral analysis and rapid method development. Our empirical results highlight BIRD-INTERACT's difficulty: GPT-5 completes only 8.67% of tasks in c-Interact and 17.00% in a-Interact. Analysis via memory grafting and Interaction Test-time Scaling validates the importance of effective interaction for complex, dynamic text-to-SQL tasks.

Moyu Zhang, Yun Chen, Yujun Jin et al.

Recently, a growing body of research has focused on either optimizing CTR model architectures to better model feature interactions or refining training objectives to aid parameter learning, thereby achieving better predictive performance. However, previous efforts have primarily focused on the training phase, largely neglecting opportunities for optimization during the inference phase. Infrequently occurring feature combinations, in particular, can degrade prediction performance, leading to unreliable or low-confidence outputs. To unlock the predictive potential of trained CTR models, we propose a Model-Agnostic Test-Time paradigm (MATT), which leverages the confidence scores of feature combinations to guide the generation of multiple inference paths, thereby mitigating the influence of low-confidence features on the final prediction. Specifically, to quantify the confidence of feature combinations, we introduce a hierarchical probabilistic hashing method to estimate the occurrence frequencies of feature combinations at various orders, which serve as their corresponding confidence scores. Then, using the confidence scores as sampling probabilities, we generate multiple instance-specific inference paths through iterative sampling and subsequently aggregate the prediction scores from multiple paths to conduct robust predictions. Finally, extensive offline experiments and online A/B tests strongly validate the compatibility and effectiveness of MATT across existing CTR models.

Shengsheng Zhang, Yaping Huang, Yun Chen et al.

Most existing Grammatical Error Correction (GEC) methods based on sequence-to-sequence mainly focus on how to generate more pseudo data to obtain better performance. Few work addresses few-shot GEC domain adaptation. In this paper, we treat different GEC domains as different GEC tasks and propose to extend meta-learning to few-shot GEC domain adaptation without using any pseudo data. We exploit a set of data-rich source domains to learn the initialization of model parameters that facilitates fast adaptation on new resource-poor target domains. We adapt GEC model to the first language (L1) of the second language learner. To evaluate the proposed method, we use nine L1s as source domains and five L1s as target domains. Experiment results on the L1 GEC domain adaptation dataset demonstrate that the proposed approach outperforms the multi-task transfer learning baseline by 0.50 $F_{0.5}$ score on average and enables us to effectively adapt to a new L1 domain with only 200 parallel sentences.

Liner Yang, Chencheng Wang, Yun Chen et al.

Due to the lack of parallel data in current Grammatical Error Correction (GEC) task, models based on Sequence to Sequence framework cannot be adequately trained to obtain higher performance. We propose two data synthesis methods which can control the error rate and the ratio of error types on synthetic data. The first approach is to corrupt each word in the monolingual corpus with a fixed probability, including replacement, insertion and deletion. Another approach is to train error generation models and further filtering the decoding results of the models. The experiments on different synthetic data show that the error rate is 40% and the ratio of error types is the same can improve the model performance better. Finally, we synthesize about 100 million data and achieve comparable performance as the state of the art, which uses twice as much data as we use.