Infographics are composite visual artifacts that combine data visualizations with textual and illustrative elements to communicate information. While recent text-to-image (T2I) models can generate aesthetically appealing images, their reliability in generating infographics remains unclear. Generated infographics may appear correct at first glance but contain easily overlooked issues, such as distorted data encoding or incorrect textual content. We present IGenBench, the first benchmark for evaluating the reliability of text-to-infographic generation, comprising 600 curated test cases spanning 30 infographic types. We design an automated evaluation framework that decomposes reliability verification into atomic yes/no questions based on a taxonomy of 10 question types. We employ multimodal large language models (MLLMs) to verify each question, yielding question-level accuracy (Q-ACC) and infographic-level accuracy (I-ACC). We comprehensively evaluate 10 state-of-the-art T2I models on IGenBench. Our systematic analysis reveals key insights for future model development: (i) a three-tier performance hierarchy with the top model achieving Q-ACC of 0.90 but I-ACC of only 0.49; (ii) data-related dimensions emerging as universal bottlenecks (e.g., Data Completeness: 0.21); and (iii) the challenge of achieving end-to-end correctness across all models.
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
The conventional Retrieval-Augmented Generation (RAG) paradigm of injecting raw retrieved texts into the Large Language Model (LLM)’s context often results in suboptimal integration of retrieved information. This paper proposes to bridge retrieval results and the LLM’s reasoning ability through Verbal Annotations, analytic narratives that explicitly articulate the logical connection between a search query and retrieved contexts. Our empirical investigation reveals the potential of Verbal Annotations to substantially enhance the LLM’s ability to generate accurate, contextually-grounded responses. Motivated by this finding, we introduce Verbal-R3, a novel agentic RAG framework that consists of a Generator and a Verbal Reranker. The Generator performs iterative retrieval and reasoning, while the Verbal Reranker returns relevance scores and Verbal Annotations to guide the reasoning and answering process of the Generator. The inference process of Verbal-R3 is further refined through relevance-guided test-time scaling, which efficiently allocates test-time compute for effective trajectory expansion. Verbal-R3 achieves state-of-the-art performance on complex Question Answering benchmarks, validating the effectiveness of the proposed framework.
Action Boundary Blindness: When LLM Agents Cannot Tell Where One Action Ends and Another Begins
PDF ↗Large language model (LLM) agents excel at multi-step tasks yet frequently exhibit Action Boundary Blindness—the inability to correctly determine action granularity, scope, and completeness. Grounded in Event Segmentation Theory from cognitive science, we formalize three violation types: granularity confusion, scope creep, and boundary ambiguity. We propose four automatic metrics—Action Boundary Score (ABS), Granularity Alignment Rate (GAR), Scope Violation Rate (SVR), and Boundary-Aware Success Rate (BASR)—requiring no human annotation. Experiments on 1,655 tasks across six benchmarks (\tau-bench, WebArena, ALFWorld, TheAgentCompany, OSWorld) with seven LLMs reveal that: (1) the best model achieves only 0.424 ABS; (2) using a multi-label attribution framework validated by inter-annotator agreement (\kappa = 0.78), boundary blindness is the primary failure mode in 37.2% of failures (25.8% as sole cause; 55.9% total involvement including contributing factors); (3) under-action dominates at 48.4%; (4) BASR is consistently \sim4 points lower than traditional success rate, exposing “lucky successes.” Critically, Explicit Boundary Prompting (EBP) improves ABS by 0.08–0.13 across all models, demonstrating that boundary blindness is better characterized as an elicitation gap rather than a fundamental capability limitation—LLMs possess latent boundary perception not activated by default. This finding has implications for alignment and instruction tuning. We validate metrics through state-based cross-validation and human audit, estimating \sim22% false positive rate from valid alternative paths, with model rankings remaining stable (Spearman \rho = 1.0).
As Multimodal Large Language Models (MLLMs) acquire stronger reasoning capabilities to handle complex, multi-image instructions, this advancement may pose new safety risks. We study this problem by introducing MIR-SafetyBench, the first benchmark focused on multi-image reasoning safety, which consists of 2,676 instances across a taxonomy of 9 multi-image relations. Our extensive evaluations on 19 MLLMs reveal a troubling trend: models with more advanced multi-image reasoning can be more vulnerable on MIR-SafetyBench. Beyond attack success rates, we find that many responses labeled as safe are superficial, often driven by misunderstanding or evasive, non-committal replies. We further observe that unsafe generations exhibit lower attention entropy than safe ones on average. This internal signature suggests a possible risk that models may over-focus on task solving while neglecting safety constraints.
Memory-Augmented Generation (MAG) extends large language models with external memory to support long-context reasoning, but existing approaches largely rely on semantic similarity over monolithic memory stores, entangling temporal, causal, and entity information. This design limits interpretability and alignment between query intent and retrieved evidence, leading to suboptimal reasoning accuracy. In this paper, we propose MAGMA, a multi-graph agentic memory architecture that represents each memory item across orthogonal semantic, temporal, causal, and entity graphs. MAGMA formulates retrieval as policy-guided traversal over these relational views, enabling query-adaptive selection and structured context construction. By decoupling memory representation from retrieval logic, MAGMA provides transparent reasoning paths and fine-grained control over retrieval. Experiments on LoCoMo and LongMemEval demonstrate that MAGMA consistently outperforms state-of-the-art agentic memory systems in long-horizon reasoning task.
ES4R: Speech Encoding Based on Prepositive Affective Modeling for Empathetic Response Generation
PDF ↗Empathetic speech dialogue requires not only understanding linguistic content but also perceiving rich paralinguistic information such as prosody, tone, and emotional intensity for affective understandings. Existing speech-to-speech large language models either rely on ASR transcription or use encoders to extract latent representations, often weakening affective information and contextual coherence in multi-turn dialogues. To address this, we propose ES4R, a framework for speech-based empathetic response generation. Our core innovation lies in explicitly modeling structured affective context before speech encoding, rather than relying on implicit learning by the encoder or explicit emotion supervision. Specifically, we introduce a dual-level attention mechanism to capture turn-level affective states and dialogue-level affective dynamics. The resulting affective representations are then integrated with textual semantics through speech-guided cross-modal attention to generate empathetic responses. For speech output, we employ energy-based strategy selection and style fusion to achieve empathetic speech synthesis. ES4R consistently outperforms strong baselines in both automatic and human evaluations and remains robust across different Large Language Model (LLM) backbones. Code: https://github.com/Bean0901/ES4R.
ARCHITECT: Uncertainty-Aware Dynamic Tool Learning via Causal Intervention for Open-World Agents
PDF ↗Dynamic tool generation empowers Large Language Model (LLM) agents to synthesize tools on demand, yet a critical challenge remains: 32.4% of generated tools fail on first invocation. We present Causal Tool Diagnosis (CTD), a principled framework that moves beyond black-box reliability prediction to interpretable failure attribution. CTD constructs a Structural Causal Model (SCM) capturing how specification quality, code characteristics, and execution environment jointly determine tool outcomes. Uniquely leveraging code’s intervenability, we conduct controlled sandbox experiments to estimate causal effects—an advantage unavailable in pure text generation. CTD jointly predicts confidence (Spearman rank correlation coefficient \rho=0.90) and root cause attribution (78% accuracy), with attributions directly guiding targeted repairs (+9.6% success rate over error-type classification). Our ARCHITECT framework, integrating CTD throughout the tool lifecycle, achieves state-of-the-art on four benchmarks including StableToolBench (+3.8%), MINT (+4.6%), T-Eval (+3.7%), and SWE-bench Lite (+2.4%), with consistent improvements across all settings.
Multilingual large language models achieve impressive cross-lingual performance despite largely monolingual pretraining. While bilingual data in pretraining corpora is widely believed to enable these abilities, details of its contributions remain unclear. We investigate this question by pretraining models from scratch under controlled conditions, comparing the standard web corpus with a monolingual-only version that removes all multilingual documents. Despite constituting only 2% of the corpus, removing bilingual data causes translation performance to drop 56% in BLEU, while behaviour on cross-lingual QA and general reasoning tasks remains stable, with training curves largely overlapping the baseline. To understand this asymmetry, we categorize bilingual data into parallel (14%), code-switching (72%), and miscellaneous documents (14%) based on the semantic relevance of content in different languages. We then conduct granular ablations by reintroducing parallel or code-switching data into the monolingual-only corpus. Our experiments reveal that parallel data almost fully restores translation performance (91% of the unfiltered baseline), whereas code-switching contributes minimally. Other cross-lingual tasks remain largely unaffected by either type. These findings reveal that translation critically depends on systematic token-level alignments from parallel data, whereas cross-lingual understanding and reasoning appear to be achievable even without bilingual data.
Rethinking Reasoning-Intensive Retrieval: Evaluating and Advancing Retrievers in Agentic Search Systems
PDF ↗Reasoning-intensive retrieval aims to surface evidence that maximizes downstream reasoning utility rather than only topical similarity. This capability is increasingly vital for agentic retriever-in-the-loop systems such as Deep-Research. However, existing retriever evaluation benchmarks, exemplified by Bright, provide narrow gold sets and evaluate retrievers in isolation, which obscures their value inside realistic agent workflows. We introduce Bright-Pro, an evaluation framework that assesses the effectiveness of retrievers in agentic search systems. Bright-Pro covers a broad range of queries across diverse professional domains. For each query, we provide expert-annotated reasoning aspects, positive documents, a reference response, and evaluation rubrics, enabling fine-grained assessment of retriever performance. Beyond static evaluation, we further assess retrievers in the context of agentic search systems, measuring their practical utility when serving as core components within agentic workflows. Using Bright-Pro, we evaluate classical lexical, general-purpose, and reasoning-intensive retrievers, providing actionable insights for future retriever development.
Joint Knowledge Base Completion and Question Answering by Combining Large Language Models and Small Language Models
PDF ↗Knowledge Bases (KBs) play a key role in various applications. As two representative KB-related tasks, knowledge base completion (KBC) and knowledge base question answering (KBQA) are closely related and inherently complementary with each other. Thus, it will be beneficial to solve the task of joint KBC and KBQA to make them reinforce each other. However, existing studies usually rely on the small language model (SLM) to enhance them jointly, and the large language model (LLM)’s strong reasoning ability is ignored. In this paper, by combining the strengths of the LLM with the SLM, we propose a novel framework JCQL, which can make these two tasks enhance each other in an iterative manner. To make KBC enhance KBQA, we augment the LLM agent-based KBQA model’s reasoning paths by incorporating an SLM-trained KBC model as an action of the agent, alleviating the LLM’s hallucination and high computational costs issue in KBQA. To make KBQA enhance KBC, we incrementally fine-tune the KBC model by leveraging KBQA’s reasoning paths as its supplementary training data, improving the ability of the SLM in KBC. Extensive experiments over two public benchmark data sets demonstrate that JCQL surpasses all baselines for both KBC and KBQA tasks.
Beyond training-time optimization, scaling test-time computation has emerged as a key paradigm to extend the reasoning capabilities of Large Language Models (LLMs). However, most existing methods adopt a rigid Planning-before-Trial (PbT) policy, which inefficiently allocates test-time compute by incurring planning overhead even on directly solvable problems. We propose Planning-after-Trial (PaT), an adaptive policy for code generation that invokes a planner only upon verification failure. This adaptive policy naturally enables a heterogeneous model configuration: a cost-efficient model handles generation attempts, while a powerful model is reserved for targeted planning interventions. Empirically, across multiple benchmarks and model families, our approach significantly advances the cost-performance Pareto frontier. Notably, our heterogeneous configuration achieves performance comparable to a large homogeneous model while reducing inference cost by approximately 69%.
Watermarking has emerged as a promising technique for tracing the authorship of content generated by large language models (LLMs). Among existing approaches, the KGW scheme is particularly attractive due to its versatility, efficiency, and effectiveness in natural language generation.However, KGW’s effectiveness degrades significantly under low-entropy settings such as code generation and mathematical reasoning. A crucial step in the KGW method is random vocabulary partitioning, which enables adjustments to token selection based on specific preferences. Our study revealed that the next-token probability distribution plays an critical role in determining how much, or even whether, we can modify token selection and, consequently, the effectiveness of watermarking.We refer to this characteristic, associated with the probability distribution of each token prediction, as watermark strength. In cases of random vocabulary partitioning, the lower bound of watermark strength is dictated by the next-token probability distribution. However, we found that, by redesigning the vocabulary partitioning algorithm, we can potentially raise this lower bound. In this paper, we propose SSG (Sort-then-Split by Groups), a method that partitions the vocabulary into two logit-balanced subsets. This design lifts the lower bound of watermark strength for each token prediction, thereby improving watermark detectability. Experiments on code generation and mathematical reasoning datasets demonstrate the effectiveness of SSG.
As large language models (LLMs) continue to advance, aligning these models with human preferences has emerged as a critical challenge. Traditional alignment methods, relying on human or LLM annotated datasets, are limited by their resource-intensive nature, inherent subjectivity, misalignment with real-world user preferences, and the risk of feedback loops that amplify model biases. To overcome these limitations, we introduce WildFeedback, a novel framework that leverages in-situ user feedback during conversations with LLMs to create preference datasets automatically. Given a corpus of multi-turn user-LLM conversation, WildFeedback identifies and classifies user feedback to LLM responses between conversation turns. The user feedback is then used to create examples of preferred and dispreferred responses according to users’ preference. Our experiments demonstrate that LLMs fine-tuned on WildFeedback dataset exhibit significantly improved alignment with user preferences, as evidenced by both traditional benchmarks and our proposed checklist-guided evaluation. By incorporating in-situ feedback from actual users, WildFeedback addresses the scalability, subjectivity, and bias challenges that plague existing approaches, marking a significant step toward developing LLMs that are more responsive to the diverse and evolving needs of their users.
Explainable diagnosis requires that authoritative medical knowledge provide the rationales linking a patient’s clinical manifestations to the diagnostic conclusion. Although large language models (LLMs) hold great potential to facilitate explainable diagnosis, their effectiveness is often constrained by insufficient diagnostic expertise. To address this limitation, we propose Self-learned Explainable Knowledge Augmented Diagnosis (SEKAD), a unified LLM-based framework for faithful and explainable diagnosis. Our approach builds a high-quality diagnostic knowledge base through a record-driven explanation learning paradigm, as well as applies this knowledge via an explanation-based diagnostic process that ensures faithful inference. Experiments on the DiReCT and JAMA benchmarks show that SEKAD consistently outperforms strong baselines across the metrics. In particular, on the DiReCT benchmark, SEKAD improves the explanation completeness metric from 64.5% to 76.9% over the best existing methods, highlighting its effectiveness in enhancing diagnostic explainability and showing that our text mining approach produces knowledge that is both reliable in quality and large in quantity.
The Dominance of Text Space: Unveiling the Asymmetric Nature of Cross-Modal Alignment in Large Language Models
PDF ↗Recent advancements in Multimodal Large Language Models (MLLMs) have largely been driven by aligning visual encoders with pre-trained Large Language Models (LLMs). While effective, the geometric nature of this alignment remains under-explored. Existing methods often assume a symmetric interaction between visual and textual modalities, implying that both spaces adapt to each other. In this paper, we challenge this assumption and propose the "Text Space as Anchor" hypothesis. We argue that the semantic space of LLMs is rigid, anisotropic, and dominant; thus, effective cross-modal alignment may be an asymmetric projection of visual features onto this pre-existing text manifold without distorting it. We identify a potential issue in current parameter-efficient tuning paradigms where task-specific visual adjustments inadvertently disrupt the projector’s geometry, leading to "catastrophic forgetting" of the alignment mechanism itself. To address this, we introduce Anchor-Preserving Projection (APP), a novel method that regularizes the projector to maintain the geometric structure of the text embedding space via spectral filtering. Extensive experiments on 8 diverse cross-modal tasks and 3 pure language benchmarks demonstrate that APP preserves the LLM’s inherent linguistic capabilities (e.g., MMLU, GSM8K) and reduces object hallucination significantly better than standard fine-tuning methods. We release our code.
Current LLM safety research predominantly focuses on mitigating **Goal Hijacking**, preventing attackers from redirecting a model’s high-level objective (e.g., from "summarizing emails" to "phishing users"). In this paper, we argue that this perspective is incomplete and highlight a critical vulnerability in **Reasoning Alignment**. We expose the inherent fragility of current alignment techniques by proposing a new adversarial prompt attack paradigm: **Reasoning Hijacking**. To demonstrate this vulnerability, we instantiate it via the **Criteria Attack**, which subverts model judgments by injecting spurious decision criteria without altering the high-level task goal. Unlike Goal Hijacking, which attempts to override the system prompt, Reasoning Hijacking keeps the task goal intact but manipulates the model’s decision-making logic by injecting spurious reasoning shortcuts. Through extensive experiments on three different tasks (toxic comment, negative review, and spam detection), we demonstrate that even state-of-the-art models are highly fragile, consistently prioritizing injected heuristic shortcuts over rigorous semantic analysis. Crucially, because the model’s explicit intent remains aligned with the user’s instructions, these attacks can bypass defenses designed to detect goal deviation (e.g., SecAlign, StruQ), revealing a fundamental blind spot in the current safety landscape. Data and code are available at [https://github.com/Yuan-Hou/criteria_attack](https://github.com/Yuan-Hou/criteria_attack).
Recent multimodal large language models (MLLMs), such as GPT-4o, Gemini 1.5/2.5 Pro, and Reka Core, have advanced audio-visual reasoning capabilities, achieving strong performance in tasks like cross-modal understanding and generation. However, our DeafTest uncovers unanticipated failures: most of the state-of-the-art MLLMs struggle with very simple audio tasks, such as distinguishing louder sounds or sound counting. This raises a fundamental question—does a deficiency in low-level audio perception constrain higher-level audio-visual reasoning? To address this, we introduce AV-Odyssey Bench—a comprehensive benchmark of 4,555 meticulously designed problems that integrate text, audio, and visual modalities. Each task requires models to unify cross-modal reasoning, leveraging synchronized audio-visual cues to infer solutions. By structuring questions as multiple-choice, we ensure objective, reproducible evaluations without reliance on subjective human or LLM-based judgments. Through comprehensive benchmarking of closed-source and open-source models, we showcase: (i) current MLLMs lack robust audio-visual integration ability and (ii) performance on DeafTest (Pearson’s r = 0.945) strongly correlates with AV-Odyssey accuracy. These findings challenge assumptions about models’ multimodal proficiency and highlight fundamental audio perception as a reasoning bottleneck. We believe that our results provide concrete guidance for future dataset design, alignment strategies, and architectures.
Mitigating Tokenization-Induced Distance Distortion in Long-Context Multilingual Machine Translation
PDF ↗Multilingual neural machine translation (MNMT) models degrade in performance as input context length increases, causing positional encoding schemes to misinterpret token distances. Existing absolute and relative positional encodings rely on fixed token indices and implicitly assume uniform semantic density, which breaks down for long-context inputs. We introduce DCARPE, a tokenization-aware adaptive positional encoding that conditions relative positional bias on input-level sequence length and fragmentation statistics, allowing the model to reinterpret positional distance when tokenization-induced inflation arises rather than semantic factors. Evaluations on JW300 and out-of-distribution FLORES-200 demonstrate consistent improvements in long-context robustness, achieving gains of up to +10.81 ChrF++ and +8.00 BLEU over baselines.
How Context Shapes Truth: Geometric Transformations of Statement-level Truth Representations in LLMs
PDF ↗Large Language Models (LLMs) often encode whether a statement is true as a vector in their residual stream activations. These vectors, also known as truth vectors, have been studied in prior work, however how they change when context is introduced remains unexplored. We study this question by measuring (1) the directional change (\theta) between the truth vectors with and without context and (2) the relative magnitude of the truth vectors upon adding context. Across four LLMs and four datasets, we find that (1) truth vectors are roughly orthogonal in early layers, converge in middle layers, and may stabilize or continue increasing in later layers; (2) adding context generally increases the truth vector magnitude, i.e., the separation between true and false representations in the activation space is amplified; (3) larger models distinguish relevant from irrelevant context mainly through directional change (\theta), while smaller models show this distinction through magnitude differences. We also find that context conflicting with parametric knowledge produces larger geometric changes than parametrically aligned context. Collectively, these findings provide a geometric characterization of how context transforms the truth vector in the activation space of LLMs.
An Existence Proof for Neural Language Models That Can Explain Garden-Path Effects via Surprisal
PDF ↗Surprisal theory hypothesizes that the difficulty of human sentence processing increases linearly with surprisal, the negative log-probability of a word given its context. Computational psycholinguistics has tested this hypothesis using language models (LMs) as proxies for human prediction. While surprisal derived from recent neural LMs generally captures human processing difficulty on naturalistic corpora that predominantly consist of simple sentences, it severely underestimates processing difficulty on sentences that require syntactic disambiguation (garden-path effects). This leads to the claim that the processing difficulty of such sentences cannot be reduced to surprisal, although it remains possible that neural LMs simply differ from humans in next-word prediction. In this paper, we investigate whether it is truly impossible to construct a neural LM that can explain garden-path effects via surprisal. Specifically, instead of evaluating off-the-shelf neural LMs, we fine-tune these LMs on garden-path sentences so as to better align surprisal-based reading-time estimates with actual human reading times. Our results show that fine-tuned LMs do not overfit and successfully capture human reading slowdowns on held-out garden-path items; they even improve predictive power for human reading times on naturalistic corpora and preserve their general LM capabilities. These results provide an existence proof for a neural LM that can explain both garden-path effects and naturalistic reading times via surprisal, but also raise a theoretical question: what kind of evidence can truly falsify surprisal theory?