VideoTIR introduces a reinforcement‑learning approach to improve long‑video understanding by encouraging multimodal large language models to use comprehensive multi‑level toolkits efficiently. It combines Zero‑RL and SFT cold‑starting strategies to help models retrieve and focus on meaningful video segments, images, and regions, thereby reducing hallucinations. The method includes Toolkit Action Grouped Policy Optimization (TAGPO) to streamline tool‑calling and a sandbox‑based trajectory synthesis framework for high‑quality data, achieving strong results on three long‑video QA benchmarks.
By Zhe Gao, Shiyu Shen, Taifeng Chai, Weinong Wang, Haotian Xu, Xing Wu, Wenbin Li, Qi Fan, Yang Gao, Dacheng Tao
arXiv:2609.08515v1 Announce Type: cross
Abstract: As Large Language Models (LLMs) are increasingly integrated into human society, aligning them with pluralistic social values has become a critical pr...
By Yuemei Xu, Kexin Xu, Jian Zhou, Haoyu Lu, Yequan Wang, Aishan Liu
arXiv:2601.20845v2 Announce Type: replace
Abstract: Time series forecasting is a fundamental problem with applications in climate, energy, healthcare, and finance. Many existing approaches require do...
By Olaf Yunus Laitinen Imanov, Derya Umut Kulali, Taner Yilmaz
Large language models are typically trained under uniform token weighting, which allows frequent and low-information tokens to dominate learning and can increase the tendency to memorize surface-level...
Uncertainty estimation for medical vision--language models (VLMs) using conformal prediction has gained increasing attention due to its distribution-free coverage guarantees. However, standard conform...
Large language models remain fragile against malicious fine-tuning, motivating training-time defenses against harmful persona drift. Preventative Steering injects undesirable-trait persona vectors dur...
Compute governance today is a governance of training: the thresholds, reporting requirements, and frontier-AI regimes now in force attach to training compute and treat the trained model as the regulat...
Modern TTS systems approach human quality for high-resource languages but degrade when clean speech data is scarce. Modern Greek exemplifies this, lacking the curated corpora behind state-of-the-art s...
Real-world detectors must often interpret functional or ambiguous prompts, yet conventional models such as YOLO remain restricted to fixed class lists. Even open-vocabulary models like YOLO-World freq...
Surveys remain the primary way researchers grasp the lineage of methods within an AI subfield, but they scale poorly against the current rate of publication. Existing taxonomy-induction methods are la...
We introduce BuzzASR, a collection of language-specialized fine-tuned Whisper models adapted for automatic speech recognition (ASR) in 102 languages. Large end-to-end Transformer-based ASR models such...
The paper investigates image tokenizers as the visual language of unified multimodal models by creating a controlled autoregressive testbed that tracks task‑specific validation losses during multimodal continual pretraining across text, image, text‑to‑image, and image‑to‑text predictions. It shows that losses must be analyzed by task, that the loss–performance relationship varies with the token space, and that better reconstruction does not always lead to stronger downstream performance. The study also demonstrates how tokenizer design choices—such as discriminator use, semantic supervision, and vocabulary size—affect joint modeling and downstream results.
The paper investigates how per-domain data composition during the mid‑training phase (between pre‑training and alignment) affects model performance. Experiments with Qwen3‑8B‑Base across five KOR‑Bench domains show that a moderate coverage band (10%‑40%) yields the best performance for each domain, and that alignment passes cannot fully close the gaps created by suboptimal mid‑training allocations. Additionally, zero coverage during mid‑training severely degrades accuracy, while a carefully tuned allocation can provide the largest overall pipeline improvement.
SeGDeP introduces a decoupled prompt system that separates semantic identification from spatial grounding in reasoning segmentation. It uses a semantic prompt branch and an independent geometric projection path to produce semantic features and a DETR‑predicted box, which jointly condition a SAM mask decoder. The model achieves high cIoU and gIoU scores while adapting only a small fraction of Qwen3‑VL parameters via LoRA.
The paper introduces Adaptive Anisotropic Attention (AAA), a method that splits self‑attention into temporal and spatial paths for axis‑structured signals like EEG. A learned gate combines the two paths for each token, and the resulting AXON model outperforms dense attention baselines on six EEG tasks and shows benefits in audio spectrogram experiments. The study demonstrates that aligning attention with natural signal axes provides a useful inductive bias.
The study investigates how different types of caregiver feedback influence grammar learning by training small GPT‑2‑style models on child‑directed language and fine‑tuning them with reinforcement learning. Four feedback categories—communicative, structural alignment, semantic contingency, and affective—were evaluated, with structural alignment showing the strongest improvement in grammaticality and communicative feedback yielding moderate gains. Semantic contingency and affective feedback did not enhance grammaticality, though they may aid other language learning aspects, indicating that various feedback forms contribute complementarily to language acquisition.
FPicker is a topology‑guided framework for filament tracing in low‑signal Cryo‑EM images. It combines a center‑endpoint representation with an open‑curve evolution module to model non‑cyclic connectivity, overcoming limitations of pixel‑wise segmenters, box‑based detectors, sequential trackers, and traditional active contours. On simulated benchmarks, FPicker improves mean spatio‑angular precision by over 40 % and reduces topological gap rates by more than 60 % under extreme noise, and it achieves state‑of‑the‑art performance on real EMPIAR data after fine‑tuning.
RefDiT is a new framework for reference-guided image generation that addresses the shortcomings of previous methods when handling complex scenes with multiple objects. It introduces local region guidance by decomposing a single identifier token into attribute-level signals, allowing the model to learn correspondences between tokens and specific regions of a reference image. The approach incorporates a low-rank adapter (LoRA) within a diffusion transformer (DiT) to adjust the inference prompt based on user-provided guidance context, thereby enabling more precise local attribute control.
By Rameshwar Mishra, Srikrishna Karanam, A V Subramanyam
The paper studies how to assess risk in intent-based networking by examining uncertainty in large language model (LLM) translations of high‑level intents into network configurations. It introduces two uncertainty signals: sampling‑based predictive uncertainty for ranking translations by risk, and token‑level entropy for locating ambiguous source tokens. Experiments on a Llama‑3.1‑8B‑Instruct model fine‑tuned for Juniper EX3300 show that predictive uncertainty can rank risky translations, while token entropy correlates with ambiguity from parameters and keywords.
By Ala' A. Alsamarneh, Omar Alhussein