While large language models (LLMs) augmented with agentic search capabilities show promise for legal reasoning, they overlook a fundamental constraint that applicable law must match the temporal context of each case, as retroactive application of statutes violates core legal principles and leads to erroneous conclusions. Our observations reveal that current legal LLMs suffer from temporal bias anchored to their training cutoff, while search agents rarely incorporate temporal constraints into queries, and that web search alone cannot provide the precise statute and precedent citations that legal reasoning demands. To address these challenges, we propose LegalSearch-R1, an end-to-end reinforcement learning framework that pairs local statute RAG for precise article matching with online web search for broader legal knowledge, trained on temporally-indexed data spanning multiple amendment periods to enforce temporal consistency. Extensive experiments on our benchmark covering 13 legal tasks demonstrate that our 7B-parameter agent outperforms state-of-the-art deep research frameworks and specialized legal LLMs by 12.9% to 29.8%, surpasses baselines by 57.7% to 80.3% on temporal consistency, and exhibits robust out-of-domain generalization. The code and data are available at https://anonymous.4open.science/r/LegalSearch-R1.

Accurately assessing model confidence is essential for deploying large language models (LLMs) in mission-critical factual domains. While retrieval-augmented generation (RAG) is widely adopted to improve grounding, confidence calibration in RAG settings remains poorly understood. We conduct a systematic study across four benchmarks, revealing that LLMs exhibit poor calibration performance especially when noisy contexts are retrieved. Specifically, contradictory or irrelevant evidence tends to exacerbate the model’s overconfidence issue. To address this, we propose *NOVA Rules* (**NO**ise-aware **V**erbal Confidence C**A**Libration Rules) to provide a principled foundation for resolving overconfidence under noise. We further design **NOVA**, a noise-aware calibration framework that synthesizes supervision from ~2K HotpotQA examples guided by these rules. By performing supervised fine-tuning (SFT) with this data, NOVA equips models with intrinsic noise awareness without relying on stronger teacher models. Empirical results show that NOVA yields substantial gains, improving ECE scores by 10.9% in-domain and 8.0% out-of-domain. By bridging the gap between retrieval noise and verbal calibration, NOVA paves the way for both accurate and epistemically reliable LLMs.

While LVLMs rapidly improve, long-video question answering still remains challenging: relevant evidence is sparse, and question-relevant context often fails to provide cues that discriminate the correct answer from plausible alternatives. Diagnostic analysis on a manually annotated subset of MMR-V shows that prior agentic systems substantially improve cue retrieval over direct VLM inference yet fail to achieve a corresponding gain in answer accuracy, indicating that the bottleneck lies in option-discriminative evidence rather than topical relevance alone. We propose PACE (Progressive Acquisition of Critical Evidence), a factor-guided framework for long-video evidence acquisition. PACE proceeds in two stages: it first indexes clip-level descriptions guided by question-derived factors without observing the candidate answers; it then uses the candidate answers to derive contrastive cues and queries the index for verification. On MMR-V with the open-source Qwen3-VL backbone, PACE achieves 42.6% accuracy, outperforming direct inference and prior agentic baselines including Deep Video Discovery (DVD). On the same diagnostic subset, PACE recovers 66.9% of the annotated cues, providing empirical evidence that its gains are associated with improved evidence recovery rather than stronger answer-side priors alone. Consistent gains over DVD on LVBench, Video-MME, EgoSchema, and LongVideoBench suggest that option-aware evidence acquisition transfers beyond MMR-V.

Session history is a common way of recording user interacting behaviors throughout a browsing activity with multiple products. For example, if an user clicks a product webpage and then leaves, it might because there are certain features that don’t satisfy the user, which serve as an important indicator of on-the-spot user preferences. However, all prior works fail to capture and model customer intention effectively because insufficient information exploitation and only apparent information like descriptions and titles are used. There is also a lack of data and corresponding benchmark for explicitly modeling intention in E-commerce product purchase sessions. To address these issues, we introduce the concept of an intention tree and propose a dataset curation pipeline. Together, we construct a sibling multimodal benchmark, SessionIntentBench, that evaluates L(V)LMs’ capability on understanding inter-session intention shift with four subtasks. With 1,952,177 intention entries, 1,132,145 session intention trajectories, and 13,003,664 available tasks mined using 10,905 sessions, we provide a scalable way to exploit the existing session data for customer intention understanding. We conduct human annotations to collect ground-truth label for a subset of collected data to form an evaluation gold set. Extensive experiments on the annotated data further confirm that current L(V)LMs fail to capture and utilize the intention across the complex session setting. Further analysis show injecting intention enhances LLMs’ performances.

Legal rules encompass not only codified statutes but also implicit adjudicatory principles derived from precedents that contain discretionary norms, social morality, and policy. While computational legal research has advanced in applying established rules to cases, inducing legal rules from judicial decisions remains understudied across jurisdictions. The advent of Large Language Models (LLMs) offers unprecedented potential to automate the extraction of such latent principles, yet progress is stymied by the absence of formal task definitions and benchmarks. To address this gap, we formalize Legal Rule Induction (LRI) as the task of deriving concise, generalizable doctrinal rules from analogous precedents, distilling their shared preconditions, normative behaviors, and legal consequences. We further propose a reproducible pipeline for LRI dataset construction and, instantiating it on Chinese law as a representative jurisdiction, introduce the first LRI benchmark comprising 5,121 case sets (38,088 court cases) for model tuning and 216 expert-annotated gold test sets. Experimental results reveal that: 1) State-of-the-art LLMs struggle with over-generalization and hallucination; 2) Training on our dataset markedly enhances LLMs’ capabilities in capturing nuanced rule patterns across similar cases.

Long-horizon LLM agents typically choose from an environment-enumerated list of grounded actions (Planning with Actions, PwA). While this interface reduces formatting risk, scaling the list increases screening load and can bottleneck planning. We study Planning with Schemas (PwS), which replaces action enumeration with a small set of action templates that the agent instantiates into executable actions. We introduce a cognitive-bandwidth perspective of this trade-off and quantify screening vs. instantiation pressure using context length proxy and failure actions proxy. Across four interactive environments, we identify an inflection point: PwA excels at small interfaces but degrades as action lists grow, whereas PwS becomes increasingly competitive and often superior. A distractor-injection stress test reproduces the PwA drop by enlarging the action space alone, isolating screening load and highlighting schema instantiation as the key remaining bottleneck.

Multimodal abductive reasoning — the generation and selection of explanatory hypotheses from partial observations — is a cornerstone of intelligence. Current evaluations of such ability in vision–language models (VLMs) are largely confined to static, single-agent tasks. Inspired by Dixit, we introduce DixitWorld, a comprehensive evaluation suite designed to deconstruct this challenge. DixitWorld features two core components: DixitArena, a dynamic, multi-agent environment that evaluates both hypothesis generation (a “storyteller” crafting cryptic clues) and hypothesis selection (“listeners” choosing the target image from decoys) under imperfect information; and DixitBench, a static QA benchmark that isolates the listener’s task for efficient, controlled evaluation. Results from DixitArena reveal distinct, role-dependent behaviors: smaller open-source models often excel as creative storytellers, producing imaginative yet less discriminative clues, whereas larger proprietary models demonstrate superior overall performance, particularly as listeners. Performance on DixitBench strongly correlates with listener results in DixitArena, validating it as a reliable proxy for hypothesis selection. Our findings reveal a key trade-off between generative creativity and discriminative understanding in multimodal abductive reasoning, a central challenge for developing more balanced and capable vision-language agents.

Large Language Model (LLM) routing is a pivotal technique for navigating a diverse landscape of LLMs, enabling the selection of the best-performing LLMs for specific user queries while balancing performance and cost. However, current routing approaches often face limitations in scalability when dealing with a large pool of specialized LLMs, or in their adaptability to extending model scope and evolving capability domains. To overcome those challenges, we propose InferenceDynamics, a flexible and scalable multi-dimensional routing framework by modeling the capability and knowledge of models. We operate it on our comprehensive dataset RouteMix, and demonstrate its effectiveness and generalizability in group-level routing using modern benchmarks including MMLU-Pro, GPQA, BigGenBench, and LiveBench, showcasing its ability to identify and leverage top-performing models for given tasks, leading to superior outcomes with cost efficiency. The broader adoption of InferenceDynamics can empower users to harness the full specialized potential of the LLM ecosystem, and our code will be made publicly available to encourage further research.

Building effective knowledge graphs (KGs) for Retrieval-Augmented Generation (RAG) is pivotal for advancing question answering (QA) systems. However, its effectiveness is hindered by a fundamental disconnect: the knowledge graph (KG) construction process is decoupled from its downstream application, yielding suboptimal graph structures. To bridge this gap, we introduce AutoGraph-R1, the first framework to directly optimize KG construction for task performance using Reinforcement Learning (RL). AutoGraph-R1 trains an LLM constructor by framing graph generation as a policy learning problem, where the reward is derived from the graph’s functional utility in a RAG pipeline. We design two novel, task-aware reward functions, one for graphs as knowledge carriers and another as knowledge indices. Across multiple QA benchmarks, AutoGraph-R1 consistently enables graph RAG methods to achieve significant performance gains over using task-agnostic baseline graphs. Our work shows it is possible to close the loop between construction and application, shifting the paradigm from building intrinsically ‘good‘ graphs to building demonstrably ‘useful‘ ones.

Conceptualization, a fundamental element of human cognition, plays a pivotal role in human generalizable reasoning.Generally speaking, it refers to the process of sequentially abstracting specific instances into higher-level concepts and then forming abstract knowledge that can be applied in unfamiliar or novel situations. This enhances models’ inferential capabilities and supports the effective transfer of knowledge across various domains.Despite its significance, the broad nature of this term has led to inconsistencies in understanding conceptualization across various works, as there exists different types of instances that can be abstracted in a wide variety of ways.There is also a lack of a systematic overview that comprehensively examines existing works on the definition, execution, and application of conceptualization to enhance reasoning tasks.In this paper, we address these gaps by first proposing a categorization of different types of conceptualizations into four levels based on the types of instances being conceptualized, in order to clarify the term and define the scope of our work.Then, we present the first comprehensive survey of over 150 papers, surveying various definitions, resources, methods, and downstream applications related to conceptualization into a unified taxonomy, with a focus on the entity and event levels.Furthermore, we shed light on potential future directions in this field and hope to garner more attention from the community.

Lateral thinking is essential in breaking away from conventional thought patterns and finding innovative solutions to problems. Despite this, language models often struggle with reasoning tasks that require lateral thinking. In this paper, we present our system for SemEval-2024 Task 9’s BrainTeaser challenge, which requires language models to answer brain teaser questions that typically involve lateral reasoning scenarios. Our framework is based on large language models and incorporates a zero-shot prompting method that integrates conceptualizations of automatically detected instances in the question. We also transform the task of question answering into a declarative format to enhance the discriminatory ability of large language models. Our zero-shot evaluation results with ChatGPT indicate that our approach outperforms baselines, including zero-shot and few-shot prompting and chain-of-thought reasoning. Additionally, our system ranks ninth on the official leaderboard, demonstrating its strong performance.

In the information era, the vast proliferation of online content poses significant challenges, particularly concerning the trustworthiness of these digital statements, which can have profound societal implications. Although it is possible to manually annotate and verify the authenticity of such content, the sheer volume and rapid pace of information generation render this approach impractical, both in terms of time and cost. Therefore, it is imperative to develop automated systems capable of validating online claims, ensuring that users can use the wealth of information available on the Internet effectively and reliably. Using primarily ChatGPT and the Google search API, GProofT fact checking framework generates question-answer pairs to systematically extract and verify the facts within claims. Based on the outcomes of these QA pairs, claims are subsequently labeled as Supported, Conflicted Evidence/Cherry-Picking, or Refuted. Shown by extensive experiments, GProofT Retrieval generally performs effectively in fact-checking and makes a substantial contribution to the task. Our code is released on https://github.com/HKUST-KnowComp/GProofT.

Improving user experience and providing personalized search results in E-commerce platforms heavily rely on understanding purchase intention. However, existing methods for acquiring large-scale intentions bank on distilling large language models with human annotation for verification. Such an approach tends to generate product-centric intentions, overlook valuable visual information from product images, and incurs high costs for scalability. To address these issues, we introduce MIND, a multimodal framework that allows Large Vision-Language Models (LVLMs) to infer purchase intentions from multimodal product metadata and prioritize human-centric ones. Using Amazon Review data, we apply MIND and create a multimodal intention knowledge base, which contains 1,264,441 intentions derived from 126,142 co-buy shopping records across 107,215 products. Extensive human evaluations demonstrate the high plausibility and typicality of our obtained intentions and validate the effectiveness of our distillation framework and filtering mechanism. Further experiments reveal the positive downstream benefits that MIND brings to intention comprehension tasks and highlight the importance of multimodal generation and role-aware filtering. Additionally, MIND shows robustness to different prompts and superior generation quality compared to previous methods.

In this paper, we present our framework for DialAM-2024 TaskA: Identification of Propositional Relations and TaskB: Identification of Illocutionary Relations. The goal of task A is to detect argumentative relations between propositions in an argumentative dialogue. i.e., Inference, Conflict, Rephrase while task B aims to detect illocutionary relations between locutions and argumentative propositions in a dialogue. e.g., Asserting, Agreeing, Arguing, Disagreeing. Noticing the definition of the relations are strict and professional under the context of IAT framework, we meticulously curate prompts which not only incorporate formal definition of the relations, but also exhibit the subtle differences between them. The PTLMs are then fine-tuned on the human-designed prompts to enhance its discrimination capability in classifying different theoretical relations by learning from the human instruction and the ground truth samples. After extensive experiments, a fine-tuned DeBERTa-v3-base model exhibits the best performance among all PTLMs with an F1 score of 78.90% on Task B. It is worth noticing that our framework ranks #2 in the ILO - General official leaderboard.

The sequential process of conceptualization and instantiation is essential to generalizable commonsense reasoning as it allows the application of existing knowledge to unfamiliar scenarios. However, existing works tend to undervalue the step of instantiation and heavilyrely on pre-built concept taxonomies and human annotations to collect both types of knowledge, resulting in a lack of instantiated knowledge to complete reasoning, high cost, and limited scalability. To tackle these challenges, we introduce CANDLE (ConceptuAlizationand INstantiation Distillation from Large Language ModEls), a distillation framework that iteratively performs contextualized conceptualization and instantiation over commonsense knowledge bases by instructing large language models to generate both types of knowledge with critic filtering. By applying CANDLE to ATOMIC (Sap et al., 2019a), we construct a comprehensive knowledge base comprising six million conceptualizations and instantiated commonsense knowledge triples. Both types of knowledge are firmly rooted in the original ATOMIC dataset, and intrinsic evaluations demonstrate their exceptional quality and diversity. Empirical results indicate that distilling CANDLE on student models provides benefits across three downstream tasks. Our data and models are publicly available at https://github.com/HKUST-KnowComp/CANDLE.

Sign Language Translation (SLT) is a complex task that involves accurately interpreting sign language gestures and translating them into spoken or written language and vice versa. Its primary objective is to facilitate communication between individuals with hearing difficulties using deep learning systems. Existing approaches leverage gloss annotations of sign language gestures to assist the model in capturing the movement and differentiating various gestures. However, constructing a large-scale gloss-annotated dataset is both expensive and impractical to cover multiple languages, and pre-trained generative models cannot be efficiently used due to the lack of textual source context in SLT. To address these challenges, we propose a gloss-free framework for the WMT23 SLT task. Our system primarily consists of a visual extractor for extracting video embeddings and a generator responsible for producing the translated text. We also employ an embedding alignment block that is trained to align the embedding space of the visual extractor with that of the generator. Despite undergoing extensive training and validation, our system consistently falls short of meeting the baseline performance. Further analysis shows that our model’s poor projection rate prevents it from learning diverse visual embeddings. Our codes and model checkpoints are available at https://github.com/HKUST-KnowComp/SLT.

In this paper, we present our system for the textual entailment identification task as a subtask of the SemEval-2023 Task 7: Multi-evidence Natural Language Inference for Clinical Trial Data. The entailment identification task aims to determine whether a medical statement affirms a valid entailment given a clinical trial premise or forms a contradiction with it. Since the task is inherently a text classification task, we propose a system that performs binary classification given a statement and its associated clinical trial. Our proposed system leverages a human-defined prompt to aggregate the information contained in the statement, section name, and clinical trials. Pre-trained language models are then finetuned on the prompted input sentences to learn to discriminate the inference relation between the statement and clinical trial. To validate our system, we conduct extensive experiments with a wide variety of pre-trained language models. Our best system is built on DeBERTa-v3-large, which achieves an F1 score of 0.764 and secures the fifth rank in the official leaderboard.Further analysis indicates that leveraging our designed prompt is effective, and our model suffers from a low recall. Our code and pre-trained models are available at [https://github.com/HKUST-KnowComp/NLI4CT](https://github.com/HKUST-KnowComp/NLI4CT).

The task of zero-shot commonsense question answering evaluates models on their capacity to reason about general scenarios beyond those presented in specific datasets. Existing approaches for tackling this task leverage external knowledge from CommonSense Knowledge Bases (CSKBs) by pre-training the model on synthetic QA pairs constructed from CSKBs. In these approaches, negative examples (distractors) are formulated by randomly sampling from CSKBs using fairly primitive keyword constraints. However, two bottlenecks limit these approaches: the inherent incompleteness of CSKBs limits the semantic coverage of synthetic QA pairs, and the lack of human annotations makes the sampled negative examples potentially uninformative and contradictory. To tackle these limitations above, we propose Conceptualization-Augmented Reasoner (CAR), a zero-shot commonsense question-answering framework that fully leverages the power of conceptualization. Specifically, CAR abstracts a commonsense knowledge triple to many higher-level instances, which increases the coverage of the CSKB and expands the ground-truth answer space, reducing the likelihood of selecting false negative distractors. Extensive experiments demonstrate that CAR more robustly generalizes to answering questions about zero-shot commonsense scenarios than existing methods, including large language models, such as GPT3.5 and ChatGPT. Our code, data, and model checkpoints are available at https://github.com/HKUST-KnowComp/CAR.

Zero-shot commonsense Question-Answering (QA) requires models to reason about general situations beyond specific benchmarks. State-of-the-art approaches fine-tune language models on QA pairs constructed from CommonSense Knowledge Bases (CSKBs) to equip the models with more commonsense knowledge in a QA context. However, current QA synthesis protocols may introduce noise from the CSKBs and generate ungrammatical questions and false negative options, which impede the model’s ability to generalize. To address these issues, we propose QADYNAMICS, a training dynamics-driven framework for QA diagnostics and refinement. Our approach analyzes the training dynamics of each QA pair at both the question level and option level, discarding machine-detectable artifacts by removing uninformative QA pairs and mislabeled or false-negative options. Extensive experiments demonstrate the effectiveness of our approach, which outperforms all baselines while using only 33% of the synthetic data, even including LLMs such as ChatGPT. Moreover, expert evaluations confirm that our framework significantly improves the quality of QA synthesis. Our code and model checkpoints are available at https://github.com/HKUST-KnowComp/QaDynamics.

A main goal of Argument Mining (AM) is to analyze an author’s stance. Unlike previous AM datasets focusing only on text, the shared task at the 10th Workshop on Argument Mining introduces a dataset including both texts and images. Importantly, these images contain both visual elements and optical characters. Our new framework, TILFA (A Unified Framework for Text, Image, and Layout Fusion in Argument Mining), is designed to handle this mixed data. It excels at not only understanding text but also detecting optical characters and recognizing layout details in images. Our model significantly outperforms existing baselines, earning our team, KnowComp, the 1st place in the leaderboard of Argumentative Stance Classification subtask in this shared task.

Commonsense reasoning, aiming at endowing machines with a human-like ability to make situational presumptions, is extremely challenging to generalize. For someone who barely knows about “meditation,” while is knowledgeable about “singing,” he can still infer that “meditation makes people relaxed” from the existing knowledge that “singing makes people relaxed” by first conceptualizing “singing” as a “relaxing event” and then instantiating that event to “meditation.”This process, known as conceptual induction and deduction, is fundamental to commonsense reasoning while lacking both labeled data and methodologies to enhance commonsense modeling. To fill such a research gap, we propose CAT (Contextualized ConceptuAlization and InsTantiation),a semi-supervised learning framework that integrates event conceptualization and instantiation to conceptualize commonsense knowledge bases at scale. Extensive experiments show that our framework achieves state-of-the-art performances on two conceptualization tasks, and the acquired abstract commonsense knowledge can significantly improve commonsense inference modeling. Our code, data, and fine-tuned models are publicly available at [https://github.com/HKUST-KnowComp/CAT](https://github.com/HKUST-KnowComp/CAT).