Recent advancements in large language models (LLMs) have empowered autonomous web agents to execute natural language instructions directly on real-world webpages. However, existing agents often struggle with complex tasks involving dynamic interactions and long-horizon execution due to rigid planning strategies and hallucination-prone reasoning. To address these limitations, we propose WebUncertainty, a novel autonomous agent framework designed to tackle dual-level uncertainty in planning and reasoning. Specifically, we design a Task Uncertainty-Driven Adaptive Planning Mechanism that adaptively selects planning modes to navigate unknown environments. Furthermore, we introduce an Action Uncertainty-Driven Monte Carlo tree search (MCTS) Reasoning Mechanism. This mechanism incorporates the Confidence-induced Action Uncertainty (ConActU) strategy to quantify both aleatoric uncertainty (AU) and epistemic uncertainty (EU), thereby optimizing the search process and guiding robust decision-making. Experimental results on the WebArena and WebVoyager benchmarks demonstrate that WebUncertainty achieves superior performance compared to state-of-the-art baselines.
论文检索
输入标题、作者或关键词，从 100,903 篇学术成果中精准定位
Code-Switching Information Retrieval: Benchmarks, Analysis, and the Limits of Current Retrievers
PDF ↗Code-switching is a pervasive linguistic phenomenon in global communication, yet modern information retrieval systems remain predominantly designed for, and evaluated within, monolingual contexts. To bridge this critical disconnect, we present a holistic study dedicated to code-switching IR. We introduce CSR-L (Code-Switching Retrieval benchmark-Lite), constructing a dataset via human annotation to capture the authentic naturalness of mixed-language queries. Our evaluation across statistical, dense, and late-interaction paradigms reveals that code-switching acts as a fundamental performance bottleneck, degrading the effectiveness of even robust multilingual models. We demonstrate that this failure stems from substantial divergence in the embedding space between pure and code-switched text. Scaling this investigation, we propose CS-MTEB, a comprehensive benchmark covering 11 diverse tasks, where we observe performance declines of up to 27%. Finally, we show that standard multilingual techniques like vocabulary expansion are insufficient to resolve these deficits completely. These findings underscore the fragility of current systems and establish code-switching as a crucial frontier for future IR optimization.
Retrieving relevant tables from extensive databases for a given natural language query is essential for accurately answering questions in tasks such as text-to-SQL. Existing table retrieval approaches select a pre-determined set of k tables with the highest similarity to the query. However, the number of required tables varies across queries and cannot be known in advance. Enforcing a fixed number of retrieved tables regardless of the query may either retrieve an undersized set, failing to obtain all necessary evidence, or retrieve an oversized pool, including irrelevant tables. To address this issue, we propose an adaptive table retrieval method that adjusts the number of tables retrieved according to the requirements of each query. Specifically, we utilize an adaptive thresholding mechanism to selectively retrieve tables and integrate a sliding-window reranking algorithm to efficiently process a large table corpus. Extensive experiments on Spider, BIRD, and Spider 2.0 demonstrate that our method effectively addresses the limitations of the top-k retrieval strategy, improving performance in retrieval and downstream tasks. Our code and data are available at https://anonymous.4open.science/r/Adaptive-Table-Retrieval.
Current evaluation methods for large language models (LLMs) primarily rely on static benchmarks, presenting two major challenges: limited knowledge coverage and fixed difficulties that mismatch with the evaluated LLMs. These limitations lead to superficial assessments of LLM knowledge, thereby impeding the targeted model optimizations.To bridge this gap, we propose JudgeAgent, a knowledge-driven and dynamic evaluation framework for LLMs.To address the challenge of limited knowledge coverage, JudgeAgent leverages LLM agents equipped with context graphs to traverse knowledge structures systematically for question generation.Furthermore, to mitigate data contamination and difficulty mismatch, it adopts a difficulty-adaptive and multi-turn interview mechanism.Thereby, JudgeAgent can achieve comprehensive evaluations and facilitate more effective improvement of LLMs.Empirical results demonstrate that JudgeAgent enables more comprehensive evaluations and facilitates effective model iterations, highlighting the potential of this knowledge-driven and dynamic evaluation paradigm.The source code is available on https://github.com/DataArcTech/JudgeAgent.
Bridging Reasoning and Action: Hybrid LLM–RL Framework for Efficient Cross-Domain Task-Oriented Dialogue
PDF ↗Cross-domain task-oriented dialogue requires reasoning over implicit and explicit feasibility constraints while planning long-horizon, multi-turn actions. Large language models (LLMs) can infer such constraints but are unreliable over long horizons, while Reinforcement learning (RL) optimizes long-horizon behavior yet cannot recover constraints from raw dialogue. Naively coupling LLMs with RL is therefore brittle: unverified or unstructured LLM outputs can corrupt state representations and misguide policy learning. Motivated by this, we propose Verified LLM-Knowledge empowered RL (VLK-RL), a hybrid framework that makes LLM-derived constraint reasoning usable for RL. VLK-RL first elicits candidate constraints with an LLM and then verifies them via a dual-role cross-examination procedure to suppress hallucinations and cross-turn inconsistencies. The verified constraints are mapped into ontology-aligned slot–value representations, yielding a structured, constraint-aware state for RL policy optimization. Experiments across multiple benchmarks demonstrate that VLK-RL significantly improves generalization and robustness, outperforming strong single-model baselines on long-horizon tasks.
Generative Retrieval (GR) has emerged as a promising text-to-image paradigm, yet it suffers from limited semantic discriminability, alignment bias, and closed-set restrictions. To address these challenges, we propose SIGMA, a novel framework for Semantic Internalization for Generative Multimodal Alignment. SIGMA constructs multi-granularity hierarchical identifiers to ensure unique, semantically consistent image representations. We further introduce a progressive semantic internalization training strategy augmented with semantic soft labels, which captures fine-grained text-image affinities and enables inductive identifier assignment for unseen samples realizing open-set dynamic indexing capabilities. Experiments on the Flickr30K and MS-COCO datasets demonstrate that SIGMA outperforms state-of-the-art baselines, achieving average Recall@1, Recall@5, and Recall@10 improvements of 10.65%, 8.50%, and 7.00%, respectively.
From Mimesis to Metamorphosis: Evolving VLM Judges via In-Context Comparing and Knowledge Internalization
PDF ↗Vision-language models (VLMs) are increasingly adopted as judges for subjective assessment, yet absolute scoring remains brittle due to inconsistent scales and inherent preference biases. To bridge this gap, we propose \text{S}^2\text{AD} (**Semantic-Anchored Scale-Agnostic Distillation**), a novel easy-to-hard framework that operationalizes subjective assessment as comparative analysis, conceptualizing the judge’s evolution from mimesis to metamorphosis. In Stage 1 (Mimesis), we introduce Dynamic Soft Positioning (DSP) to train the judge to compare a query against retrieved reference images, establishing a relative evaluation space that ensures consistent ordering under heterogeneous scales. In Stage 2 (Metamorphosis), this comparative capability is internalized via Language Buttons—discrete semantic levels serving as a retrieval-free internal reference. Optimized with Group Relative Policy Optimization (GRPO), \text{S}^2\text{AD} achieves efficient, scale-steerable inference that adapts to diverse grading standards. Our framework reaches state-of-the-art performance across multiple benchmarks, validating the effectiveness of internalized comparative priors for robust, rank-invariant, and scale-steerable evaluation. The code is available at: https://github.com/SpatialVision-Research/SSAD_ACL2026_Findings.
Large language model-based agents operating in long-horizon interactions require memory systems that support temporal consistency, multi-hop reasoning, and evidence-grounded reuse across sessions. Existing approaches largely rely on unstructured retrieval or coarse abstractions, which often lead to temporal conflicts, brittle reasoning, and limited traceability. We propose MemWeaver, a unified memory framework that consolidates long-term agent experiences into three interconnected components: a temporally grounded graph memory for structured relational reasoning, an experience memory that abstracts recurring interaction patterns from repeated observations, and a passage memory that preserves original textual evidence. MemWeaver employs a dual-channel retrieval strategy that jointly retrieves structured knowledge and supporting evidence to construct compact yet information-dense contexts for reasoning. Experiments on the LoCoMo benchmark demonstrate that MemWeaver substantially improves multi-hop and temporal reasoning accuracy while reducing input context length by over 95% compared to long-context baselines.
AI agents operating on user interfaces must understand how interfaces communicate state and feedback to act reliably. As a core communicative modality, animations are increasingly used in modern interfaces, serving critical functional purposes beyond mere aesthetics. Thus, understanding UI animation is essential for comprehensive interface interpretation. However, recent studies of Vision Language Models (VLMs) for UI understanding have focused primarily on static screenshots, leaving it unclear how well these models handle dynamic UI animations. To address this gap, we created AniMINT, a novel dataset of 300 densely annotated UI animation videos.We systematically evaluate state-of-the-art VLMs on UI animation understanding, including their abilities to perceive the animation effects, identify animation purposes, and interpret animation meaning. Our results show that VLMs can reliably detect primitive motion.However, their high-level animation interpretation remains inconsistent, with substantial gaps relative to human performance. Finally, we use Motion, Context, and Perceptual Cues (MCPC) to probe factors affecting VLM performance, revealing key bottlenecks and directions for future improvement.
Multimodal large language models (MLLMs) are increasingly deployed in Web-scale applications—such as image search, social media captioning, and e-commerce product description generation—where factual consistency is critical for user trust and content reliability. However, we observe that MLLMs frequently hallucinate in these settings due to two relevant phenomena: the massive activation phenomenon and positional information decay. The former refers to the tendency of attention mechanisms to concentrate on a small set of tokens with extreme activation values in query and key projections, which play indispensable roles in contextual understanding. In our investigation, we discover that perturbing these tokens leads to significant performance drops, highlighting their utmost importance. As for positional information decay, it occurs due to the common rotary position encoding strategy, where the attention to early visual tokens diminishes over time, especially in long-sequence generation tasks, such as image caption. To address these challenges, we propose TokenTruth, a token-level intervention strategy that dynamically suppresses irrelevant visual tokens while preserving key contextual signals. Our method is grounded in an in-depth analysis of massive activations and attention sink behaviors, and introduces a targeted token penalty mechanism that reallocates attention more faithfully toward informative visual regions. Extensive experiments demonstrate that TokenTruth significantly improves factual consistency across various MLLMs on standard image understanding benchmarks.
Medical coding converts free-text clinical notes into standardized diagnostic and procedural codes, which are essential for billing, hospital operations, and medical research. Unlike ordinary text classification, it requires multi-step reasoning: extracting diagnostic concepts, applying guideline constraints, mapping to hierarchical codebooks, and ensuring cross-document consistency. Recent advances leverage agentic LLMs, but most rely on rigid, manually crafted workflows that fail to capture the nuance and variability of real-world documentation, leaving open the question of how to systematically learn effective workflows. We present MedDCR, a closed-loop framework that treats workflow design as a learning problem. A Designer proposes workflows, a Coder executes them, and a Reflector evaluates predictions and provides constructive feedback, while a memory archive preserves prior designs for reuse and iterative refinement. On benchmark datasets, MedDCR outperforms state-of-the-art baselines and produces interpretable, adaptable workflows that better reflect real coding practice, improving both the reliability and trustworthiness of automated systems.
Long-reasoning models achieve strong accuracy on complex reasoning tasks, but their extended reasoning trajectories incur substantial memory and latency costs. Several existing shortening methods rely on additional supervision or multi-stage post-training, which primarily reduces inference length and does not reduce the rollout tokens during on-policy reinforcement learning (RL). We instead target on-policy response shortening, aiming to improve both inference efficiency and RL training throughput. However, because on-policy RL couples optimization with exploration, naively penalizing length can destabilize training and suppress exploration. To impose length pressure safely, we propose a lazy length penalty integrated into the rule-based RL pipeline: it activates only on correct trajectories, only after training accuracy enters a stably improving regime, and only when responses exceed a tolerance band beyond the minimal correct length. Across four settings, our method significantly reduces response length without extra training stages while maintaining or improving performance. In a logic reasoning setting, we achieve a 40% reduction in step-averaged response length alongside a 14-point gain in performance. For math problems, we reduce step-averaged response length by 33% while preserving performance.
Large language models are often adapted using parameter-efficient techniques such as Low-Rank Adaptation (LoRA), formulated as y = W_0x + BAx, where W_0 is the pre-trained parameters and x is the input to the adapted layer. While multi-adapter extensions often employ multiple LoRAs, prior studies suggest that the inner A matrices are highly similar during training and thus suitable for sharing. We revisit this phenomenon and find that this similarity is largely attributable to the identical initialization rather than shared knowledge, with B playing a more critical role in knowledge encoding and transfer. Motivated by these insights, we propose **ALoRA**, an asymmetric multi-LoRA design with multiple A matrices and a single shared B in multi-task fine-tuning, and **Fed-ALoRA**, which shares B across clients in federated fine-tuning under both homogeneous and heterogeneous settings, through a novel matrix decomposition strategy to accommodate heterogeneous ranks across clients. Experiments on commonsense reasoning, math reasoning, multi-task NLP dataset, and federated NLP dataset demonstrate that our methods achieve more balanced performance across tasks with comparable or superior average accuracy relative to existing multi-LoRA approaches. The code is available at https://github.com/OptMN-Lab/ALoRA.
Identifying Collective Intelligence Factor in LLM Agent Groups for Generalizable Multi-Agent System Design
PDF ↗Large language model (LLM)-based multi-agent systems (MASs) have shown impressive performance in solving a wide range of complex problems. However, previous studies mainly focus on designing customized MAS for specific tasks, while a critical research problem remains unclear: Do LLM agent groups exhibit a form of “general intelligence” that reflects their general ability across various tasks? Researchers have found a Collective Intelligence (CI) factor in human groups that captures their general capability. Inspired by this, in this study, we aim to investigate whether an analogous CI factor also exists in LLM agent groups, which is crucial for building generalizable MAS. Motivated by human cognitive psychology experiments, we construct 108 LLM agent groups with diverse group sizes, LLM compositions, and communication topologies. We systematically evaluate these groups across a wide range of tasks and analyze their performances. Our results demonstrate that an Artificial Collective Intelligence (ACI) factor can be extracted from LLM agent groups to predict the generalization performance on new tasks. Inspired by this, we train a model to predict the ACI based on the features of MAS, and show that it can be used as a plug-in to enhance the generalization ability of MAS optimization methods.
Emotional Quotient (EQ) has emerged as a competency for seamless human-AI integration. However, since traditional EQ scales focus on self-healing, directly migrating them to Large Language Models (LLMs) often leads to ignorance of healing others. While EQ metrics specifically designed for LLMs have been proposed, they remain mired in two dilemmas: dimensional deficiency and fragmented testing. Hence, this paper establishes a Quad-in-One architecture for a closed-loop EQ evaluation. First, we propose the PACE Taxonomy to define four dimensions of LLM EQ. Upon this, the Causal-PACE framework is developed to eliminate causal confounding bias triggered by the interactions among EQ dimensions, ensuring a rigorous quantification of composite EQ scores. To operationalize this framework, we implement the PACE-AB, a mutil-agent EQevaluation board system. Finally, we curate the PACE-2700 dataset, featuring 2,700 high-quality instructions, to serve as a comprehensive benchmark for large-scale validation.Experimental results demonstrate that the EQ values derived via the Causal-PACE achieve a high alignment of 89.31% with human preferences, while the automated PACE-AB system maintains a robust consistency of 83.6%. Our data is publicly available at https://anonymous.4open.science/r/PACE-2700-8E52.
The reasoning capabilities of large language models (LLMs) have improved substantially through increased test-time computation, typically in the form of intermediate tokens known as chain-of-thought (CoT). However, CoT often becomes unnecessarily long, increasing computation costs without improving accuracy and sometimes even degrading performance, a phenomenon known as “overthinking”. We propose a multi-stage efficient reasoning method that combines supervised fine-tuning—via rejection sampling or reasoning trace reformatting—with reinforcement learning using an adaptive length penalty. We introduce a lightweight reward function that penalizes tokens generated after the first correct answer, encouraging the model to perform self-verification only when beneficial. We conduct a holistic evaluation across seven diverse reasoning tasks, analyzing the accuracy–response length trade-off. Our approach reduces response length by an average of 28% for 8B models and 40% for 32B models, while incurring only minor performance drops of 1.6 and 2.5 points, respectively. Despite its conceptual simplicity, it achieves a better trade-off than more complex state-of-the-art efficient reasoning methods, scoring 76.6 on the area under the Overthinking-Adjusted Accuracy curve (\text{AUC}_{\text{OAA}})—5 points above the base model and 2.5 points above the second-best approach.
As Large Language Models (LLMs) scale to support context windows exceeding one million tokens, the linear growth of Key-Value (KV) cache imposes severe memory capacity and bandwidth bottlenecks, constraining the efficiency of long-context inference. Existing compression approaches typically prioritize tokens based on local saliency metrics to decouple prefill computation from decoding memory. However, these methods often rely on local saliency snapshots at a specific layer, thereby systematically discarding tokens that act as global information hubs across the network depth but appear temporarily dormant at the specific layer selected for pruning. To address this limitation, we propose StructKV, a structure-aware KV cache compression framework that introduces three core innovations: First, Global In-Degree Centrality aggregates attention patterns across the network depth to identify global information hubs. Second, Dynamic Pivot Detection utilizes information-theoretic metrics to adaptively locate the optimal layer for compression. Finally, Structural Propagation Decoupling separates the computational budget from the memory storage budget. Experimental results on the LongBench and RULER benchmarks demonstrate that StructKV effectively preserves long-range dependencies and retrieval robustness.
Using LLMs not to predict plans but to formalize an environment into the Planning Domain Definition Language (PDDL) has been shown to improve performance and control. While most existing methodology only applies to fully observable environments, we adapt to the more realistic and challenging partially observable environments without sufficient information to make a complete plan. We propose PDDLego+, a framework to iteratively formalize, plan, grow, and refine PDDL representations by decomposing the environment and the goal into fully observable episodes. Without fine-tuning, in-context exemplars, or trajectories, PDDLego+ improves planning success and exhibits robustness against problem complexity compared to end-to-end approaches. We also show that the domain knowledge captured after a successful trial can benefit future tasks.
Feedback to Reasoning: LLM-Assisted Molecular Optimization with Domain Feedback and Historical Reasoning
PDF ↗The success of large language models (LLMs) across domains highlights their potential in scientific tasks, with molecular optimization being a promising frontier. Traditionally, this optimization relies on iterative expert feedback to refine molecules toward desired properties, a process well aligned with LLMs’ strengths. **As an experience-driven task, molecular optimization depends critically on the domain feedback and accumulation of historical knowledge. However, none of the existing methods fully leverages such feedback and historical knowledge with reasoning traces and chemical insights.** In this work, we propose F2R: Feedback to Reasoning, a conversational molecular optimization pipeline that enables LLMs to accumulate and retrieve past actions, rationales, and feedback. Like humans, LLMs can generate imperfect reasoning; F2R is the first framework to use detailed domain feedback to critique and improve this reasoning. This transforms LLMs from passive text generators into agentic experts that learn both actions and reasoning from experience. Consequently, F2R shows remarkable performance.
Failure makes the agent stronger: Enhancing Accuracy through Structured Reflection for Reliable Tool Interactions
PDF ↗Tool-augmented large language models (LLMs) are typically trained via supervised imitation learning or coarse-grained reinforcement learning, approaches that primarily optimize one-shot tool calls. Existing practices of self-reflection largely rely on heuristic prompting or unidirectional reasoning traces: the model is encouraged to “think more,” rather than to treat error diagnosis and correction as a learnable capability. This makes them fragile in multi-turn interaction settings—once a call fails, the model tends to repeat the same mistake instead of recovering. To address this issue, we propose structured reflection, which transforms the “from error to repair” process into a first-class, controllable, and trainable action. The agent produces a concise yet precise reflection process: specifically, the model diagnoses the error based on evidence from the previous step and then proposes a correct and executable follow-up call. During training, we combine DAPO and GSPO’s objective functions and design a more principled reward mechanism tailored to tool calling, optimizing the stepwise strategy Reflect \to Call \to Final. To evaluate this capability, we introduce Tool-Reflection-Bench, a lightweight benchmark dataset that programmatically verifies structural validity, executability, parameter correctness, and result consistency. Tasks in the benchmark are constructed as miniature trajectories of Erroneous Call \to Reflection \to Corrected Call and are split into disjoint training and testing sets. Experiments on BFCL v3 and Tool-Reflection-Bench show that our method achieves significant improvements in multi-turn tool-call success rates and error recovery, while also reducing redundant calls. These results demonstrate that making reflection explicit and treating it as an optimization objective can substantially enhance the reliability of tool interaction, providing a reproducible pathway for agents to grow stronger by learning from failure. We will release all the code and datasets as open source once the paper is accepted by the community.