Tool-augmented visual reasoning systems, exemplified by Visual Programming (VP), decompose complex visual questions into executable programs that invoke specialized visual models as tools, offering interpretable and compositional reasoning. However, these systems face a fundamental optimization challenge: the non-differentiable pipeline and the absence of intermediate supervision prevent end-to-end gradient-based optimization of invoked tools using only task-level labels. Existing methods either optimize the Large Language Model (LLM) planner while neglecting tool quality, or rely on teacher models and additional data collection, none achieving true end-to-end differentiable optimization. We propose **eDoT**, the first **e**nd-to-end **D**ifferentiable **o**ptimization method for **T**ool-augmented visual reasoning. Our key insight is that any program’s variable dependency structure is isomorphic to a directed probabilistic graph. Applying Variable Elimination - an exact probabilistic inference technique - transforms the non-differentiable VP execution into a fully differentiable process, enabling gradient-based optimization with only task-level supervision, without additional data or teacher models. Experiments on GQA, NLVRv2 and Open Images demonstrate improvements of 3.7%, 7.6% and 20%, validating eDoT’s effectiveness across multiple visual models and cross-task generalization.

Retrieval shapes how language models access and cite knowledge in retrieval-augmented generation (RAG). In historical research, the goal is often to locate the exact record for a specific regnal month, where temporal alignment matters as much as topical relevance. This is especially challenging for Classical Chinese annals: time is encoded in terse, implicit, non-Gregorian reign phrases that are context-dependent, so semantically plausible evidence can still be temporally invalid. We introduce ChunQiuTR, a time-keyed retrieval benchmark built from the Spring and Autumn Annals and its exegetical tradition. It organizes records by month-level reign keys and includes chrono-near confounders that mimic real retrieval failures. We propose CTD (Calendrical Temporal Dual-encoder), a time-aware dual-encoder combining Fourier-based absolute context with relative offset biasing. Experiments show consistent gains over semantic dual-encoder baselines under time-keyed evaluation. We will release ChunQiuTR and code after the anonymity period.

Multi-round Vision-Language Model (VLM) Multi-Agent Systems (MAS) offer powerful reasoning capabilities but suffer from prohibitive costs due to static panel designs, where all N agents communicate at every T round. This approach is fundamentally inefficient, as it ignores the context-dependent and diminishing marginal utility of specific agents. To address this, we propose Nash-CredMAS, an economic framework that transforms agent selection into a dynamic resource allocation game. Unlike heuristic routing or one-time pruning, our method operates in two phases: (1) Offline Causal Value Learning, where we employ a doubly-robust (AIPW) estimator to train a context-aware value function from biased interaction logs, effectively learning the true marginal contribution of agents; and (2) Online Dynamic Auctions, where agents bid for communication slots based on their predicted utility. We formulate the inference-time selection as a submodular maximization problem under budget constraints, theoretically guaranteeing a (1 - 1/e)-approximation of the optimal coalition via a greedy strategy. Empirically, Nash-CredMAS achieves state-of-the-art results on challenging benchmarks, including MMMU and V*-Bench, while reducing token consumption by over 25% compared to static baselines. The system naturally converges to an economic equilibrium where agents actively remain silent when their marginal value does not justify the cost.

Masked Discrete Diffusion Models (MDMs) enable parallel generation via iterative refinement. However, we identify a critical decisional mismatch. The MDM architecture is inherently dynamic and capable of sensing context shifts. In contrast, prevailing decoding paradigms remain static and myopic. They treat each denoising step as an isolated snapshot, effectively discarding valuable temporal feedback that signals logical conflicts. To bridge this gap, we propose Regret-Aware Confidence Calibration (RACC). This training-free framework aligns decoding decisions with the model’s latent self-correction capabilities. RACC introduces a momentum anchor to track confidence trajectories. When a token’s probability drops abruptly below its historical trend, the system triggers a “regret” signal. Unlike expensive re-masking or lookahead search, RACC utilizes this signal to proactively demote unstable candidates. Extensive experiments on reasoning benchmarks, such as HumanEval and GSM8K, demonstrate that RACC significantly improves generation consistency. Crucially, RACC achieves these gains with zero additional inference overhead, effectively balancing decoding quality and efficiency.

Document Layout Analysis (DLA) pipelines provide structured page representations for retrieval-augmented generation, long-document question answering, and related applications. Yet their robustness evaluation remains largely area-centric. We identify this *Footprint Bias* and propose **ProSA**, a lightweight output-level auditing framework that decouples controlled probing, policy-driven targeting, and structure-aware diagnosis. ProSA combines Block-level Structural Loss Rate (**B-SLR**), granularity-aware exposure descriptors, and pathway attribution to analyze where structural identity is lost, at what exposure granularity failures emerge, and how failures propagate. Across MinerU and PP-StructureV3 on 1,000 pages, affected area weakly tracks perturbation-induced OCR instability (R² = 0.384/0.110), whereas B-SLR aligns much more closely with it (R² = 0.727/0.916). Exposure descriptors further separate occlusion- and topology-dominant pathways, while matched-footprint structural probes cause much larger downstream QA/retrieval drops than area-matched erasure. These results shift DLA robustness evaluation from footprint-based measurement toward structure-aware vulnerability auditing.

Narrow-domain finetuning makes vision-language models (VLMs) sound like domain experts while quietly making them less attentive to the image. We attribute this gap to a persistent domain prior: an input-agnostic shift in hidden state, installed by narrow finetuning, that biases the model toward domain-style outputs regardless of visual evidence. To test the claim we introduce ActDiff, a white-box diagnostic that isolates the finetuning trace by differencing base and finetuned hidden states under progressively degraded images. The trace is compact and persists under blank and Gaussian-noise inputs; the finetuned model is correspondingly less, not more, disrupted when the image is destroyed (gap 0.323 ± 0.036 nats on BIOMED, generic eval). Subtracting the trace at inference recovers 54% of the MMStar accuracy collapse on BIOMED, with a peak-and-collapse dose response that random and wrong-domain controls do not produce. A single first-moment regulariser added at training time shrinks the trace 34x and improves generic NLL. The prior thus has a clean diagnostic, a causal handle at inference, and a one-line mitigation at training.

Hybrid offline–online reinforcement learning (O2O RL) promises both sample efficiency and robust exploration, but suffers from instability due to distribution shift between offline and online data. We introduce RLPD-GX, a framework that decouples policy optimization from safety enforcement: a reward-seeking learner explores freely, while a projection-based guardian guarantees rule-consistent execution and safe value backups. This design preserves the exploratory value of online interactions without collapsing to conservative policies. To further stabilize training, we propose dynamic curricula that gradually extend temporal horizons and anneal offline–online data mixing. We prove convergence via a contraction property of the guarded Bellman operator, and empirically show state-of-the-art performance on Atari-100k, achieving a normalized mean score of 3.02 (+45% over prior hybrid methods) with stronger safety and stability. Beyond Atari, ablations demonstrate consistent gains across safety-critical and long-horizon tasks, underscoring the generality of our design. Extensive and comprehensive results highlight decoupled safety enforcement as a simple yet principled route to robust O2O RL, suggesting a broader paradigm for reconciling exploration and safety in reinforcement learning.

Diffusion Large Language Models (dLLMs) have emerged as a promising non-autoregressive paradigm for text generation, offering parallel decoding and bidirectional context modeling. However, aligning dLLMs with reinforcement learning (RL) remains a significant challenge, as the marginal likelihood of sequences in masked diffusion is typically intractable, rendering standard policy gradient methods unstable or computationally prohibitive. In this work, we propose Diffusion-Gibbs Alignment (DGA), a novel variational framework that reformulates RL for dLLMs as a distribution matching problem. DGA bypasses the explicit computation of log-probabilities by leveraging a learned energy function to model the relative quality of samples. The optimization is decoupled into two stable steps: (1) contrastive energy ranking to capture global reward structures, and (2) weighted diffusion alignment to update the policy via importance sampling. Empirically, DGA establishes a new state-of-the-art across logical reasoning (Sudoku, Countdown), mathematical reasoning (GSM8K, Math500), and code generation (HumanEval, MBPP) benchmarks. DGA offers a novel variational perspective for dLLM alignment, achieving better performance while simultaneously enhancing training speed and memory efficiency.

Vision-Language-Action (VLA) models have demonstrated impressive capabilities in generalized robotic control; however, they remain notoriously brittle to linguistic perturbations. We identify a critical "modality collapse” phenomenon where strong visual priors overwhelm sparse linguistic signals, causing agents to overfit to specific instruction phrasings while ignoring the underlying semantic intent. To address this, we propose Residual Semantic Steering (RSS), a probabilistic framework that disentangles physical affordance from semantic execution. RSS introduces two theoretical innovations: (1) Monte Carlo Syntactic Integration, which approximates the true semantic posterior via dense, LLM-driven distributional expansion, and (2) Residual Affordance Steering, a dual-stream decoding mechanism that explicitly isolates the causal influence of language by subtracting the visual affordance prior. Theoretical analysis suggests that RSS effectively maximizes the mutual information between action and intent while suppressing visual distractors. Empirical results across diverse manipulation benchmarks demonstrate that RSS achieves state-of-the-art robustness, maintaining performance even under adversarial linguistic perturbations.

This paper introduces OSC (Orchestrating Cognitive Synergy), a knowledge-aware adaptive collaboration framework designed to enhance cognitive synergy in multi-agent systems with large language models. While prior work has advanced agent selection and result aggregation, efficient linguistic interactions for deep collaboration among expert agents remain a critical bottleneck. OSC addresses this gap as a pivotal intermediate layer between selection and aggregation, introducing Collaborator Knowledge Models (CKM) to enable each agent to dynamically perceive its collaborators’ cognitive states. Through real-time cognitive gap analysis, agents adaptively adjust communication behaviors, including content focus, detail level, and expression style, using learned strategies. Experiments on complex reasoning and problem-solving benchmarks demonstrate that OSC significantly improves task performance and communication efficiency, transforming “parallel-working individuals” into a “deeply collaborative cognitive team”.

Multi-label classification (MLC) faces persistent challenges from label imbalance, spurious correlations, and distribution shifts, especially in rare label prediction. We propose the Causal Cooperative Game (CCG) framework, which models MLC as a multi-player cooperative process. CCG integrates explicit causal discovery via Neural Structural Equation Models, a counterfactual curiosity reward to guide robust feature learning, and a causal invariance loss to ensure generalization across environments, along with targeted rare label enhancement. Extensive experiments on benchmark datasets demonstrate that CCG significantly improves rare label prediction and overall robustness compared to strong baselines. Ablation and qualitative analyses further validate the effectiveness and interpretability of each component. Our work highlights the promise of combining causal inference and cooperative game theory for more robust and interpretable multi-label learning.

We introduce Fourier Domain Adapter (FDA), a novel and parameter-efficient framework for fine-tuning large-scale pre-trained language models. FDA reparameterizes the core projection operation of the adapter module directly in the Fourier domain. This involves transforming the input features via discrete Fourier transform (DFT), applying sparse learnable complex modulations in frequency space, and then back-transforming via inverse DFT, supplemented by highly compact auxiliary linear layers. This approach significantly reduces the number of trainable parameters while enhancing the model’s ability to capture salient frequency-based semantic information. Comprehensive experiments on GLUE, E2E NLG, and instruction tuning benchmarks show that our FDA consistently outperforms existing parameter-efficient fine-tuning (PEFT) methods. It can achieve better performance with nearly 100x fewer training parameters than traditional fine-tuning methods such as LoRA and AdapterH. Our results demonstrate that FDA is a robust and efficient solution for developing efficient and powerful language models.

This paper introduces DrDiff, a novel framework for long-text generation that overcomes the efficiency-quality trade-off through three core technologies. First, we design a dynamic expert scheduling mechanism that intelligently allocates computational resources during the diffusion process based on text complexity, enabling more efficient handling of text generation tasks of varying difficulty. Second, we introduce a Hierarchical Sparse Attention (HSA) mechanism that adaptively adjusts attention patterns according to a variety of input lengths, reducing computational complexity from O(n²) to O(n) while maintaining model performance. Finally, we propose a Semantic Anchor States (SAS) module that combines with DPM-solver++ to reduce diffusion steps, significantly improving generation speed. Comprehensive experiments on various long-text generation benchmarks demonstrate the superiority of our DrDiff over the existing SOTA methods.

Neural module networks (NMN) are a popular approach for grounding visual referring expressions. Prior implementations of NMN use pre-defined and fixed textual inputs in their module instantiation. This necessitates a large number of modules as they lack the ability to share weights and exploit associations between similar textual contexts (e.g. “dark cube on the left” vs. “black cube on the left”). In this work, we address these limitations and evaluate the impact of contextual clues in improving the performance of NMN models. First, we address the problem of fixed textual inputs by parameterizing the module arguments. This substantially reduce the number of modules in NMN by up to 75% without any loss in performance. Next we propose a method to contextualize our parameterized model to enhance the module’s capacity in exploiting the visiolinguistic associations. Our model outperforms the state-of-the-art NMN model on CLEVR-Ref+ dataset with +8.1% improvement in accuracy on the single-referent test set and +4.3% on the full test set. Additionally, we demonstrate that contextualization provides +11.2% and +1.7% improvements in accuracy over prior NMN models on CLOSURE and NLVR2. We further evaluate the impact of our contextualization by constructing a contrast set for CLEVR-Ref+, which we call CC-Ref+. We significantly outperform the baselines by as much as +10.4% absolute accuracy on CC-Ref+, illustrating the generalization skills of our approach.