Zixin Song

Jiazheng Kang, Bowen Zhang, Zixin Song et al.

ReAct-style agents for search-intensive, multi-step reasoning tasks rely largely on their own internal judgment to decide what evidence to seek, which reasoning or action step to take next, and when to stop, often producing shallow, redundant, or poorly targeted trajectories. Prior work has explored rubrics as external quality signals, but existing uses are mostly evaluative rather than action-guiding: rubrics typically serve as training-time rewards or post-hoc evaluators of completed outputs, and in deep-research settings they are often coarse-grained and report-level rather than step-level. We introduce Co-ReAct, a rubric-guided action-selection framework that uses rubrics as step-level guidance during inference. At each decision step, Co-ReAct injects a rubric into the agent's context to guide the next Reason-or-Act decision, specifying what the agent should target in evidence seeking, search, reasoning, or self-evaluation. To make this guidance reliable, we train a dedicated rubric generator with GRPO. Unlike prior pairwise or binary preference formulations, our objective optimizes a list-wise Spearman rank-correlation reward against multi-judge expert consensus rankings, encouraging rubrics that are discriminative rather than merely plausible. On DeepResearchBench and SQA-CS-V2, Co-ReAct consistently improves over ReAct and representative test-time compute baselines across search agents built on both 8B/14B open-source and frontier closed-source base models. The trained rubric generator can also serve as a drop-in component that improves these baselines without changing their underlying decision mechanisms. Our code is publicly available at https://github.com/ZBWpro/Co-ReAct.

Jiangwang Chen, Bowen Zhang, Zixin Song et al.

Although large language model (LLM) conversational systems process millions of multi-turn dialogues daily, they remain fundamentally reactive: they respond only after the user types a query. A key step toward proactive interaction is next-query prediction, which anticipates the user's subsequent query based solely on the preceding dialogue. Progress on this task is hindered by the lack of dedicated benchmarks and a fundamental efficiency--quality trade-off: naively concatenating full dialogue history incurs linearly growing token consumption, while truncating to the latest turn discards crucial cross-turn context. Our key insight is that accurate prediction does not require re-reading raw history; it suffices to track the user's evolving intent trajectory across topics, unresolved needs, and interest shifts. We propose OnePred, which maintains a recursively updated memory as its sole cross-turn context, bounding the per-turn cost independently of conversation length. We train the model via a two-stage reinforcement learning pipeline that first teaches what to predict, then what to compress, shaping the memory into a prediction-oriented intent chain. To establish a rigorous testbed, we introduce NQP-Bench, spanning three diverse subsets. Experiments demonstrate that OnePred reduces per-turn token consumption by up to 22$\times$ compared to full-history inputs while consistently exceeding all baselines in prediction quality, with larger gains on longer conversations. Our code is publicly available at https://github.com/ZBWpro/OnePred.

Xiaojie Hong, Zixin Song, Liangzhi Li et al.

Medical Visual Question Answering (Med-VQA) is a very important task in healthcare industry, which answers a natural language question with a medical image. Existing VQA techniques in information systems can be directly applied to solving the task. However, they often suffer from (i) the data insufficient problem, which makes it difficult to train the state of the arts (SOTAs) for the domain-specific task, and (ii) the reproducibility problem, that many existing models have not been thoroughly evaluated in a unified experimental setup. To address these issues, this paper develops a Benchmark Evaluation SysTem for Medical Visual Question Answering, denoted by BESTMVQA. Given self-collected clinical data, our system provides a useful tool for users to automatically build Med-VQA datasets, which helps overcoming the data insufficient problem. Users also can conveniently select a wide spectrum of SOTA models from our model library to perform a comprehensive empirical study. With simple configurations, our system automatically trains and evaluates the selected models over a benchmark dataset, and reports the comprehensive results for users to develop new techniques or perform medical practice. Limitations of existing work are overcome (i) by the data generation tool, which automatically constructs new datasets from unstructured clinical data, and (ii) by evaluating SOTAs on benchmark datasets in a unified experimental setup. The demonstration video of our system can be found at https://youtu.be/QkEeFlu1x4A. Our code and data will be available soon.

Bowen Zhang, Zixin Song, Chunping Li

As a fundamental task in Information Retrieval and Computational Linguistics, sentence representation has profound implications for a wide range of practical applications such as text clustering, content analysis, question-answering systems, and web search. Recent advances in pre-trained language models (PLMs) have driven remarkable progress in this field, particularly through unsupervised embedding derivation methods centered on discriminative PLMs like BERT. However, due to time and computational constraints, few efforts have attempted to integrate unsupervised sentence representation with generative PLMs, which typically possess much larger parameter sizes. Given that state-of-the-art models in both academia and industry are predominantly based on generative architectures, there is a pressing need for an efficient unsupervised text representation framework tailored to decoder-only PLMs. To address this concern, we propose CSE-SFP, an innovative method that exploits the structural characteristics of generative models. Compared to existing strategies, CSE-SFP requires only a single forward pass to perform effective unsupervised contrastive learning. Rigorous experimentation demonstrates that CSE-SFP not only produces higher-quality embeddings but also significantly reduces both training time and memory consumption. Furthermore, we introduce two ratio metrics that jointly assess alignment and uniformity, thereby providing a more robust means for evaluating the semantic spatial properties of encoding models.

Zixin Song, Bowen Zhang, Qian-Wen Zhang et al.

Conditional Semantic Textual Similarity (C-STS) measures the semantic proximity between text segments under a specific condition, thereby overcoming the ambiguity inherent in traditional STS. However, existing methods are largely confined to discriminative models, failing to fully integrate recent breakthroughs in the NLP community concerning Large Language Models (LLMs) and Reinforcement Learning (RL). RL is a particularly well-suited paradigm for this task, as it can directly optimize the non-differentiable Spearman ranking metric and guide the reasoning process required by C-STS. However, we find that naively applying listwise RL fails to produce meaningful improvements, as the model is overwhelmed by complex, coarse-grained reward signals. To address this challenge, we introduce PoLi-RL, a novel Point-to-List Reinforcement Learning framework. PoLi-RL employs a two-stage curriculum: it first trains the model with simple pointwise rewards to establish fundamental scoring capabilities, then transitions to a hybrid reward that combines pointwise, pairwise, and listwise objectives to refine the model's ability to discern subtle semantic distinctions. Crucially, we propose an innovative Parallel Slice Ranking Reward (PSRR) mechanism that computes ranking rewards in parallel slices, where each slice comprises same-indexed completions from different samples. This provides a precise, differentiated learning signal for each individual completion, enabling granular credit assignment and effective optimization. On the official C-STS benchmark, PoLi-RL achieves a Spearman correlation coefficient of 48.18, establishing a new SOTA for the cross-encoder architecture. As the first work to successfully apply RL to C-STS, our study introduces a powerful and precise paradigm for training LLMs on complex, ranking-based conditional judgment tasks.

Bowen Zhang, Zixin Song, Chunquan Chen et al.

Learning unified text embeddings that excel across diverse downstream tasks is a central goal in representation learning, yet negative transfer remains a persistent obstacle. This challenge is particularly pronounced when jointly training a single encoder for Information Retrieval (IR) and Semantic Textual Similarity (STS), two essential but fundamentally disparate tasks for which naive co-training typically yields steep performance trade-offs. We argue that resolving this conflict requires systematically decoupling task-specific learning signals throughout the training pipeline. To this end, we introduce CoDiEmb, a unified framework that reconciles the divergent requirements of IR and STS in a collaborative yet distinct manner. CoDiEmb integrates three key innovations for effective joint optimization: (1) Task-specialized objectives paired with a dynamic sampler that forms single-task batches and balances per-task updates, thereby preventing gradient interference. For IR, we employ a contrastive loss with multiple positives and hard negatives, augmented by cross-device sampling. For STS, we adopt order-aware objectives that directly optimize correlation and ranking consistency. (2) A delta-guided model fusion strategy that computes fine-grained merging weights for checkpoints by analyzing each parameter's deviation from its pre-trained initialization, proving more effective than traditional Model Soups. (3) An efficient, single-stage training pipeline that is simple to implement and converges stably. Extensive experiments on 15 standard IR and STS benchmarks across three base encoders validate CoDiEmb. Our results and analysis demonstrate that the framework not only mitigates cross-task trade-offs but also measurably improves the geometric properties of the embedding space.