研究動向 (arXiv)

機械学習・自然言語処理・コンピュータビジョンの最新プレプリント。

cs.CLcs.AIcs.LG2026年9月1日

Context-Grounding Gains Are Mediated by Pre-existing Machinery: Auditing GRPO, SFT, and DPO

Prakhar Gupta, Vaibhav Gupta

Language models can ignore prompt evidence when it conflicts with memorized knowledge. Post-training can make models follow such evidence more reliably, but it is unclear whether these gains require new machinery or strengthen machinery already present. We compare nine post-training arms spanning GRPO, SFT, and DPO from one starting checkpoint, with key comparisons extended across scales and families. We estimate a grounding direction from that checkpoint before training. Across five tested GRPO…

arXiv → PDF

cs.CVcs.AI2026年9月1日

Beyond the Image Plane: World-Grounded Queries for Multi-Object Tracking

Orcun Cetintas, Guillem Brasó, Tim Meinhardt, Laura Leal-Taixé

Monocular videos record 3D scenes as sequences of 2D image-plane projections, obscuring depth and spatial relationships. Multi-object trackers localize and associate objects primarily using appearance and geometry observed only in the image plane, inheriting these ambiguities. To address this limitation, we introduce PLANET, an end-to-end multi-object tracker designed to move beyond the image plane. As an enabling step, we lift existing 2D tracking datasets into 3D. We then form world-grounded q…

arXiv → PDF

cs.CV2026年9月1日

On-the-Fly3R: Towards Robust Online 3D Reconstruction with Feed-Forward 3R Models for Large-Scale UAV Scenarios

Zhe Shen, Liyuan Lou, Yifei Yu, Guanbo Wang, Quanjian Ji, Xin Wang, Zongqian Zhan

While feed-forward 3D reconstruction (3R) offers efficient end-to-end modeling, its application in large-scale UAV mapping is hindered by the prohibitive memory cost of Transformer attention. Current scalable streaming 3R methods assume temporally and spatially continuous inputs, rendering them ineffective for the weakly ordered or unordered image streams common in cross-strip UAV operations. To address this, we propose On-the-Fly3R, a training-free, progressive online 3D reconstruction framewor…

arXiv → PDF

cs.AIcs.CL2026年9月1日

VIBE-Bench: Evaluating Personalized Large Language Models When Profiles Don't Mean Preferences

Yiwen Jiang, Yang Deng, Stephanie Fong, Zimu Wang, Yaling Shen, Wei Feng, Hongxi Yang, Xiangyu Zhao et al.

Personalized Large Language Models (PLLMs) aim to tailor responses to individual users, where a central challenge is preference reasoning: inferring query-relevant preferences from user-related history. Existing benchmarks, however, largely assume that such preference can be retrieved from semantically related history. We study an underexplored but practically important regime, profile-preference conceptual misalignment (PRCM), where observable profile cues and query-specific preferences lie in…

arXiv → PDF

cs.ROcs.CV2026年9月1日

VerNav: Verifier-First Low-Latency Vision-and-Language Navigation

Zhixin Wang, Chengzheyi Yao, Leyuan Liu, Xiaosong Zhang, Yongzhao Zhang

Vision-and-Language Navigation (VLN) requires an agent to navigate through unseen 3D environments according to natural-language instructions. Explicit reasoning can improve instruction understanding and semantic grounding, but autoregressive generation at every step accumulates large decision-stage latency over multi-step navigation. We propose VerNav, a verifier-first framework for low-latency LLM-based VLN. The verifier reduces decision-stage latency by replacing per-step autoregressive genera…

arXiv → PDF

cs.AIcs.CL2026年9月1日

RPCBench: A Benchmark for Proactive Premise Critique in LLM-based Recommendation

Zhongru Chen, Yuan Wu, Yi Chang

Large language models are increasingly used as interactive recommender assistants. Their evaluation should therefore go beyond plausible item recommendation and test whether they can recognize flawed recommendation requests. Existing recommender benchmarks mainly assess ranking, generation, or preference satisfaction, while existing error-detection benchmarks are usually not grounded in recommendation-specific user and candidate evidence. To address this gap, we introduce RPCBench, a benchmark f…

arXiv → PDF

cs.CV2026年9月1日

A multicenter benchmark and clinically structured metric for coronary CTA report generation

Zhiyu Ye, Yue Sun, Limiao Zou, Cheng Xu, Keting Xu, Tong Hu, Yue Yu, Hairong Zheng et al.

Reliable evaluation of automated coronary computed tomography angiography (CCTA) report generation requires standardized multicentre benchmarks and clinically structured metrics. We established a four-centre benchmark comprising 3,021 CCTA series from 818 patient-report pairs to evaluate seven open-source three-dimensional vision-language models. We developed CSM$_{\text{CCTA}}$, a clinically structured metric for CCTA report evaluation, with patient-, vessel-, and segment-level variables define…

arXiv → PDF

cs.LGstat.COstat.ML2026年9月1日

When Metropolis and Hastings Meet Bradley and Terry: Exact MCMC From Preference Voting

Ariel Smogorghevski, Nir Rosenfeld, Yaniv Romano

Sampling from distributions conditioned on desired semantic properties is an emerging challenge in modern generative modeling. Metropolis-Hastings (MH) provides a principled route to conditional sampling, but requires access to exact pointwise target-density evaluations, which are not available in generative settings. Meanwhile, pairwise comparisons by humans or model "judge" are highly accessible and have proved valuable across diverse applications. We introduce Pref-MH, a general exact MH samp…

arXiv → PDF

cs.AI2026年9月1日

In-Context Neurofeedback: Can LLMs Control Their Internal Representations through Privileged Access?

Koshiro Aoki, Ryota Takatsuki, Gouki Minegishi, Yusuke Haruki, Daisuke Kawahara

Whether large language models (LLMs) can control their own internal representations matters for both machine metacognition and AI safety. A recent study applied neurofeedback to LLMs and claimed that they can control their internal representations. However, the reported control may rely on superficial mechanisms rather than genuine internal access because the control targets in that study are not privileged, meaning that a third party can infer them from the prompt. We redesign the neurofeedback…

arXiv → PDF

cs.CV2026年9月1日

HELIOS: From midnight to noon, continuous outdoor urban scene relighting

Hala Djeghim, Nathan Piasco, Luis Roldão, Moussab Bennehar, Dzmitry Tsishkou, Céline Loscos, Désiré Sidibé

Modifying the illumination of driving images is a fundamental challenge, as most datasets are captured at specific times of day. Existing methods rely on synthetic data or paired multi-illumination supervision, which limits their generalization to the diverse and challenging conditions of real-world scenarios. To address this, we propose HELIOS, a novel image relighting approach that relies on unlabeled real-world datasets without requiring any paired images for training. Our approach integrates…

arXiv → PDF