The paper introduces Evo-Rec, a three‑stage framework that improves generative recommendation by learning better reasoning traces for Semantic ID (SID) generation. First, it aligns SIDs with textual and behavioral contexts; second, it samples and selects reasoning traces that improve ground‑truth item prediction via supervised fine‑tuning; third, it refines the reasoning policy with reinforcement learning using catalog‑constrained generation and ranking‑aware feedback. Experiments on three Amazon Review datasets show Evo‑Rec consistently outperforms discriminative, generative, and other reasoning‑enhanced recommenders across all metrics.
S2Planner is a trajectory planner for autonomous driving that fuses data from three front-facing cameras, ego‑motion history, and the current driving command. It uses a fine‑tuned DINOv3 backbone with a Spatial Tuning Adapter to generate multi‑scale image features, which are then refined by a coarse‑to‑fine decoder employing trajectory self‑attention and camera‑projected cross‑attention. The key contribution lies in integrating ego‑conditioned trajectory initialization with iterative, geometry‑guided sampling of multi‑scale image features, rather than introducing a new visual backbone or attention operator.
The paper introduces a pipeline that localizes frames from historical PTZ maritime video onto a reference panorama and uses context-aware sampling to build compact, scene‑specific training sets for ship detection. By enriching frames with weather and solar metadata and applying diversity sampling, the method reduces 40,718 candidate frames to just 220 for annotation, achieving a 99.5% reduction. Fine‑tuned YOLOv26‑m on this curated subset attains high detection performance (AP50 ≈ 94.8% and AP50‑95 ≈ 75.1%).
HarnessPAI introduces a model‑ and embodiment‑agnostic harness framework for Physical AI that treats code as an executable, evolvable interface organizing action primitives. The framework operates on two timescales: within a rollout it executes an open‑loop program, and across rollouts it evolves a closed‑loop program using feedback to refine the program and distill reusable skills. Across diverse robotic platforms, HarnessPAI outperforms pure action models and code‑as‑policy baselines, achieving significant gains on tasks such as LIBERO‑PRO and RoboCasa, and enabling efficient expert‑data collection for further fine‑tuning.
The paper introduces αp-LoRA, a rank-allocation strategy for low-rank adaptation (LoRA) in large language models that uses π-regularization (0 < p < 1) to induce sparsity in rank-one components. By regularizing the energy of each component, redundant parts are encouraged to vanish while important ones are retained, and the authors derive a proximal subproblem that reduces the matrix optimization to a two‑dimensional thresholding criterion. Experiments on natural language understanding and question‑answering tasks show that αp-LoRA achieves performance competitive with existing LoRA baselines.
The paper introduces Median Temporal Ensembling, a training‑free aggregation method for action‑chunked visuomotor policies that replaces the standard exponential weighted mean with a coordinate‑wise median. This approach remains robust against adversarial corruption, maintaining a high recovery rate even as attack strength increases, and performs at least as well as the mean across numerous configurations while improving in many cases. It also handles non‑adversarial failures such as blank camera frames and shows limited impact on clean data, though it cannot counteract uniform shifts applied to all predictions.
By Yuhang Jiang
ScoutNeRV introduces a content‑adaptive initialization framework that speeds up the training of hierarchical grid‑based video implicit neural representations (INRs). By using a lightweight scout network to select a pre‑trained expert from a memory bank, the method transfers the expert’s grid and decoder parameters to a target HiNeRV model, achieving a 21.25 dB PSNR boost before fine‑tuning and reaching near‑baseline performance after only 37 epochs. The approach delivers a 9.25× wall‑clock speedup while maintaining competitive rate–distortion performance.
By Naser Alizada, Farhang Baghban, Hashem Pishkar, Ali Mousavi
Fine‑tuning large language models on parallel data can improve translation quality but also causes catastrophic forgetting of general capabilities. The study evaluates several forgetting‑mitigation methods—anchored to auxiliary data, model outputs, and base model parameters—using Llama 3.2 1B Instruct and Llama 3.1 8B Instruct on Arabic‑English and Spanish‑English translation tasks. Elastic Weight Consolidation best preserves general benchmark performance, yet only data mixing with control‑task examples maintains instruction‑following abilities such as formality and grammatical gender control, though these gains do not generalize to unseen prompts.
By Niklas Scholz, David Thulke, Abdallah Nasir, Will Allred, Evgeny Matusov, Hermann Ney
ASCIIBench is a new benchmark that evaluates large language models on generating and classifying ASCII-text images, using a dataset of 5,315 labeled ASCII images. The authors also release a fine‑tuned CLIP model adapted to capture ASCII structure for evaluation. Their analysis shows that cosine similarity on CLIP embeddings fails to separate most categories, indicating a representation bottleneck rather than generational variance.
By Kerry Luo, Michael Fu, Joshua Peguero, Husnain Malik, Anvay Patil, Joyce Lin, Megan Van Overborg, Ryan Sarmiento, Kevin Zhu
Mizar is a 159.3‑million‑parameter audio‑language model designed for devices with limited memory and computation. It couples a compact CED‑Small audio encoder with SmolLM2‑135M via a frequency‑merging mapper and is trained in three stages—audio‑language alignment, audio‑dependent fine‑tuning, and post‑training—to improve performance on audio‑question tasks. Across five random seeds, Mizar outperforms all other sub‑200M‑parameter ALMs on MMAU, MMAR, and ADQA‑clean, achieving mean accuracies of 52.92%, 42.42%, and 36.02% respectively, while enabling local inference on a single CPU with an average latency of 1.09 seconds for MMAU questions.
By Kaiyang Li, Shaobo Han, Yue Tian, Shihao Ji
The paper introduces ICM, an Intra-Class Mixing Consistency framework for unsupervised domain adaptation in semantic segmentation under adverse weather. ICM mixes regions within the same image and semantic class to maintain realistic layouts, contrasting prior methods that combine across images or domains. On the Cityscapes → ACDC benchmark, ICM achieves 75.7% mIoU, surpassing previous state‑of‑the‑art results by 1.9 percentage points.
By Boying Li, Chang Liu, Britta Ayano Wilde, Gy\"orgy Kov\'acs, Tosin Adewumi, Bj\"orn Backe, Hamam Mokayed
CasCVS‑Net is a staged multi‑task cascade that jointly performs object detection, semantic segmentation, and Critical View of Safety (CVS) assessment for laparoscopic cholecystectomy. The model couples tasks through predicted anatomy—boxes guide segmentation and masks provide region‑level features for CVS classification—allowing CVS assessment to rely solely on model predictions. Trained on the Endoscapes dataset, CasCVS‑Net outperforms state‑of‑the‑art methods, achieving higher mAP and mIoU scores across detection, segmentation, and CVS tasks, especially for rare hepatocystic structures.
By Bock-Zien Toh, Yuanchuan Ren, Tay Aw Yu, Ng Khee Ong, Zhehua Mao, Sophia Bano
This paper presents a zero‑shot framework for removing objects from real images using a frozen Stable Diffusion model, avoiding any task‑specific training. The pipeline combines SAM‑based mask construction, BLIP caption conditioning, DDIM inversion, background‑weighted masked null‑text optimization, decoder self‑attention masking, hard outside‑mask latent anchoring, and localized renoise‑denoise refinement. Experiments show effective removal of objects and context‑consistent replacement, with background‑weighted NTI especially helpful for complex backgrounds and repeated refinement reducing residual artifacts.
By Arman Taghizadeh (Institute of Cognitive Science, Osnabr\"uck University, Osnabr\"uck, Germany), Ulf Krumnack (Institute of Cognitive Science, Osnabr\"uck University, Osnabr\"uck, Germany), Kai-Uwe K\"uhnberger (Institute of Cognitive Science, Osnabr\"uck University, Osnabr\"uck, Germany)
The study evaluates whether learned context planning can outperform strong retrieval methods in long-context multiple-choice question answering. Using 503 LongBench-v2 MCQ questions and a Qwen2.5-7B-Instruct model, the planner—trained on outcome-selected traces—achieves lower accuracy than anchored hybrid retrieval and BM25 across various character budgets. Even with tight budgets, the planner only marginally improves or matches retrieval, indicating that learned planning provides a weak relevance signal rather than a replacement for robust retrieval.
By Yingrui Li, Han Chen
SpeechLLMs used in professional settings often undergo domain customisation through prompts or fine‑tuning, which can inadvertently cause the model to transcribe phonetically similar words from its context or training data, leaking private information. The authors systematically investigate this overlooked privacy risk, creating benchmarks to measure leakage rates for both prompting and fine‑tuning, and find that both mechanisms cause measurable leakage that compounds when combined. They evaluate a prompt‑level mitigation strategy and analyse the accuracy‑leakage trade‑off, concluding that fine‑tuning without context prompts offers the best balance between performance and privacy.
By Maike Z\"ufle, Jan Niehues
The paper introduces a trilingual spoken hallucination detection benchmark covering English, Russian, and Kazakh news, with 12,013 samples that include synthetic alterations and severity levels, as well as 290 fact‑checked misinformation items. Detectors are evaluated in a reference‑free setting on text, ASR transcripts, and audio, revealing that most models underperform baseline classifiers, except Gemma‑3n on transcripts. Synthetic‑trained detectors achieve high macro‑F1 scores on real‑world misinformation, but Russian provenance analysis highlights model‑dependent signals that confound synthetic benchmarks.
By Meruyert Aristombayeva, Jason S. Lucas, Chaewan Chun, Dongwon Lee
ZoomDiff is a high‑fidelity diffusion model designed to improve dual‑camera smooth zooming by producing photo‑realistic transitions. It strengthens dual‑image conditional guidance during denoising, injects flow‑aligned multi‑scale features from the VAE encoder into the decoder to recover high‑frequency details, and uses flow‑guided temporal consistency supervision to ensure smoother transitions. Experiments on synthetic and real‑world datasets show that ZoomDiff outperforms state‑of‑the‑art methods both quantitatively and qualitatively.
By Jiayi Zhang, Renlong Wu, Yukang Ding, Sibin Deng, Wangmeng Zuo
The paper investigates how large language models (LLMs) interpret negation across different logical semantics—open‑world vs. closed‑world, two‑ vs. three‑valued, and credulous vs. skeptical reasoning. Using the newly introduced NAFBench, a procedural generator that creates solver‑certified logic programs and their natural‑language verbalizations, the authors evaluate LLMs on four semantic viewpoints (SLDNF, well‑founded semantics, and stable‑model semantics). Results show a persistent gap: even the strongest models achieve only 59–74% accuracy, with many models sensitive to rule ordering and prone to overcommitment on undefined cases, though some frontier models reach near‑perfect performance on a fixed‑complexity set.
"whyItMatters":"The study highlights that current LLMs struggle to reliably follow explicitly specified negation semantics, underscoring a limitation in their logical reasoning capabilities."
By Qiming Bao, Agnieszka Mensfelt, Michael J. Witbrock, Kostas Stathis
The paper introduces SGCA‑Net, a framework for place recognition using automotive spinning FMCW radar. It combines rotation‑robust feature extraction with Spatially Gated Correlation Aggregation, which learns spatial weights to mitigate unstable radar regions while preserving informative pairwise correlations. Experiments on the MulRan dataset show SGCA‑Net outperforms state‑of‑the‑art methods in various environments, and tests on the HeRCULES dataset demonstrate its ability to generalize to unseen settings and sensors without fine‑tuning.
By Saimunur Rahman, Sagun Singh Shrestha, Abdelwahed Khamis, Peyman Moghadam
The paper introduces the Agent-Editing World Model (AEWM), a new approach that models how reasoning and actions influence future task progress instead of simulating tool responses. AEWM includes an Action Judge that classifies decisions as Critical, Exploratory, or Noisy, and a State Revision mechanism that edits noisy reasoning–action continuations from the same observed history. The integrated system, EditAct, directly updates the underlying state during real execution, leading to significant performance gains across multiple benchmarks and agent backbones.
By Shuang Sun, Guoxin Chen, Fanzhe Meng, Jia Deng, Huatong Song, Jinhao Jiang, Wayne Xin Zhao, Hongteng Xu, Ji-Rong Wen