Vision-language models, speech and cross-modal systems that read, look and listen in the same forward pass.
arXiv:2609.36440v1 Announce Type: new Abstract: Large vision-language models (LVLMs) have recently achieved remarkable progress across multimodal tasks, yet object hallucination remains a persistent...
arXiv:2609.36064v1 Announce Type: cross Abstract: Foundation models are powerful generators, but many engineering domains require structured representations that general-purpose systems handle poorly...
arXiv:2609.37043v1 Announce Type: new Abstract: Sampling from unnormalized distributions over large discrete state spaces becomes difficult when a multimodal target is far from a tractable reference....
arXiv:2609.37582v1 Announce Type: cross Abstract: Federated knowledge must remain usable by recipients with different modalities, private architectures, and tasks. We present FedSocket, which makes r...
Vision‑Language Models (VLMs) used in autonomous driving are evaluated on their ability to understand velocity through three tasks: surrounding‑agent speed, current ego speed, and short‑horizon future ego‑speed. Experiments on nuScenes show that while current ego speed can be accessed internally, verbal outputs are poor and surrounding‑agent speed is weakly encoded; temporal cues and frame order are largely ignored. Specialized driving models (Alpamayo‑1.5) improve latent speed representations but still leave gaps between internal knowledge and verbal readout, indicating that good planning outputs do not guarantee reliable dynamic state recovery.
The paper introduces Q-TOFC, a query‑guided task‑oriented visual feature compression method that uses residual vector quantization to encode merged features as compact codebook index sequences. By incorporating query relevance into feature aggregation and adding a quantization error compensation adapter, Q‑TOFC reduces visual payload by 53.6% compared to previous TOFC while preserving task performance. Experiments across seven multimodal benchmarks and latency tests confirm its effectiveness under bandwidth‑constrained uplinks.
EviViT is a lightweight attachment for pretrained vision transformers that learns where to focus detail in high‑resolution images. It uses human visual‑search traces to supervise a question‑conditioned evidence density, guiding regional re‑reading and efficient visual token allocation. The method connects regional features to the global scene via a sparse, coordinate‑aware bridge, improving fine‑grained accuracy across nine host models while using fewer tokens than global‑only processing.
The paper benchmarks vision‑language models (VLMs) on two key tasks in connectomics: synapse detection (presence and polarity) and proofreading (split and merge errors). It evaluates 21 models (19 open, 2 closed) under zero‑shot, few‑shot, and LoRA fine‑tuning, comparing them to specialist models on datasets built from public resources. While most VLMs perform at chance in zero‑shot settings, LoRA fine‑tuning with a few thousand labels brings open models to specialist performance, and adapted VLMs outperform specialists on unseen species for merge‑error detection.
MemEvo presents an automated approach to discover memory mechanisms for streaming video understanding. It frames memory design as a search over executable programs expressed in a lightweight domain-specific language, and uses a pretrained large language model to generate and refine candidate memory programs. The framework evaluates each candidate deterministically while keeping the underlying vision-language model frozen, ultimately producing a training-free, bounded-memory mechanism that shows strong performance on StreamingBench and OVO-Bench.
MatToolBench is a new benchmark that evaluates multimodal GUI agents on professional materials science software. It contains 204 tasks across 10 tools in three modalities—GUI operation, OriginPro scripting, and code-based database queries—executed inside a Windows 11 VM. The benchmark offers fine-grained, expert-decomposed scoring and a high-performing multimodal judge for aesthetic assessment, revealing that strong general benchmark performance does not transfer to scientific workflows.
The paper investigates a phenomenon called inertia in computer‑use agents, where agents repeat ineffective actions despite recognizing their futility. By analyzing high‑dimensional activation states, the authors find that inertia corresponds to an absorbing region in activation space where values become stale. They propose a method called R$^3$—Reset, Reroute, Restore—to temporarily reset the agent’s context trajectory, escape the absorbing region, and then restore historical context, achieving a 17‑55% reduction in measured inertia across models.
VISTA (Value-Informed Semantic Trust Arbitration) is a learned seven-field appraisal interface that conditions modality arbitration on concerns, event relations, and expression conditions while retaining a joint-evidence residual. It uses a log-odds decomposition to separate emotion expectation from cue diagnosticity, allowing appraisal to change how evidence is interpreted. With a shared Qwen2.5-Omni-7B backbone, VISTA achieves 64.5% conflict accuracy on CA-MER, improving on modality gating by 2.5 percentage points on conflict and 0.2 on consistency, and a frozen-backbone probe reaches 0.600 macro CCC for appraisal readout versus 0.505 for emotion-only fine-tuning.
SeekVLN is a new framework for Vision‑Language Navigation that addresses the problem of agents acting on insufficient evidence, termed Progress Myopia. It combines semantic progress reasoning with active evidence seeking, trained first with Future‑guided Reverse Generation to augment expert trajectories, and then refined via Counterfactual Contrastive Policy Optimization to reward beneficial seeking actions. Experiments on simulated benchmarks show significant gains, improving success rates by 12.7% on R2R‑CE and 7.5% on RxR‑CE, and real‑world tests demonstrate human‑like evidence‑seeking behavior.
The paper introduces CALIBRA, a calibration-first multimodal temporal learning framework designed for transferable asthma‑risk forecasting across varying patient cohorts and sensor ecosystems. It processes multiple data streams—environmental, pulmonary, symptom, medication, wearable, and context—using dedicated recurrent encoders, a reliability‑conditioned gate, and gradient‑reversal training to mitigate cohort bias. CALIBRA employs a shrinkage‑based hierarchical logistic layer for probability calibration and split conformal prediction for abstention‑capable prediction sets, achieving competitive performance on a semi‑synthetic three‑cohort benchmark with controlled distribution shift.
The paper investigates how the modality gap— the separation between image and text representations in contrastive vision‑language models—affects different downstream tasks. By showing that a single dominant direction accounts for most of the image‑text mean separation, the authors explain why reducing or removing this gap can improve zero‑shot classification, degrade retrieval, or restore performance depending on the task. The study provides a geometric framework that clarifies when and why gap interventions should be applied in vision‑language systems.
The paper investigates how multimodal large language models are more susceptible to prompt injection when adversarial instructions are presented as text rather than as non-textual inputs like images. It proposes a training‑free defense that renders untrusted payloads into typographic images (or audio) before they reach the model, a method called Pictionary. Experiments on ten models and two benchmarks show that this approach significantly lowers attack success rates while maintaining normal functionality, and that fine‑tuning on image‑rendered instructions can further reduce the modality gap.
The paper introduces a new multimodal scaling law that accounts for how the proportion of paired data influences loss in multimodal classification tasks. It shows that varying the pairing ratio under fixed total data budgets dramatically affects loss curves, with synergy only reducing when a critical threshold of paired data is reached. The proposed law decomposes total loss into four power-law components—redundancy, unique modality channels, and synergy—providing a more accurate fit (3.2% error) than existing models.
The paper introduces MATE, a reinforcement‑learning‑based post‑training framework for unified multimodal models that lets the generation and understanding branches challenge each other instead of cooperating. In MATE, each branch proposes candidate outputs that the other must reproduce, and the solver is trained on the worst‑handled candidate, creating an evolving adversarial loop without a separate adversary. Experiments on Janus‑Pro‑1B show that MATE improves generation and understanding metrics, including GenEval (+2.4), DPG‑Bench (+1.7), and an average of nine understanding benchmarks (+0.7), while enhancing consistency across image‑text cycles.
AdaKerNet is a task‑adaptive neural kernel decoder that operates on frozen multimodal representations from large foundation models, without requiring access to the models’ parameters. It learns Lipschitz‑controlled multimodal features, a reference kernel providing a soft structural prior, and a lightweight nonlinear predictor that deforms this structure. Experiments on four multimodal large language models and diverse input modalities show consistent improvements over baseline decoders, achieving up to 41% error reduction in scarce‑label settings.
The study compares human and vision‑language model (VLM) responses to cross‑modal association tasks, using identical stimuli (a pseudo‑word and two images) and recording both choices and eye movements. While larger VLMs show some alignment with human choices, their attention patterns correlate poorly with human gaze, performing no better than a simple center‑bias baseline. Fine‑tuning VLMs on human choices improves choice alignment but not attention alignment, and training on human gaze improves attention correlation without affecting choice accuracy.