Using pretrained diffusion models for Bayesian inpainting of calorimeter data, validated with posterior diagnostics.

Himanshu Raj, Roli Esha

This survey reviews decoding methods for LLMs and LVLMs, emphasizing contrastive, guided, and parallel paradigms that improve quality, control, and efficiency.

Haoran Wang, Xiongxiao Xu, Philip S. Yu et al.

Introduced verifier skipping, reducing calls by 13.5% on HumanEval while maintaining pass@1 performance.

Haoxuan Luo, Jameson Sandler, Ferdinando Fioretto

Proposes a task-relevant state transfer framework using predictive sufficiency, with fixed-length bit bounds for session handover.

Masahiro Kato, Taka Kato

Active inference-based EFE path planning enables autonomous robots to efficiently map and locate high-value regions under resource constraints, outperforming traditional methods.

Ajith Anil Meera, Pablo Lanillos, Wouter Kouw

Wyvern employs a multi-agent framework combining retrieval, generation, and claim verification, achieving 87% improvement in figure informativeness.

Beatrice Alessandra Motetti, Emilien Guandalino, Daniele Jahier Pagliari et al.

Proposes online inference in distributional TD learning, establishing weak convergence to Gaussian processes and bootstrap consistency for smooth and nonsmooth functionals.

Yang Peng, Liangyu Zhang

Multi-layer twisted yarn capacitive sensors achieve a sensitivity of 0.1331 MPa$^{-1}$ at 100kHz, with proximity detection ranges of 40-60mm, enabling tunable pressure and near-field sensing.

Ishtia Zahir, Eslam Saleh, Maryam Rezayati et al.

CoRun leverages position invariance to achieve deterministic, high-throughput LLM inference, boosting throughput by 15-324%.

Shiju Zhao, Jiacheng Yang, Qihang Chen et al.

Introduces DHD protocol to measure trajectory value of multi-agent messages, revealing that wrong messages can be helpful; over 40% of such messages aid reasoning in benchmarks.

Chih-Hsuan Yang, Anjir Ahmed Chowdhury, Cheng-Hau Yang et al.

ScienceFlow uses recoverable executable states and ESTRA for long-horizon autonomous research, boosting search efficiency.

Mingming Zhao, Jiqian Dong, Kangping Xu et al.

Mobius-v0 separates knowledge and reasoning, achieving nearly 4x inference speedup while maintaining accuracy.

Kai Chen, Jifeng Ding, Ning Ding et al.

Pairton uses an iterative reconstruction framework to reconstruct short-lived particles in high-energy collisions, achieving state-of-the-art performance on fully hadronic t¯t decays.

Andreas Hermansen, Chris Scheulen, Tobias Golling

SimpleOPD enables cross-tokenizer long-context reasoning transfer, boosting mathematical proof performance by 21.2 points on ProofBench.

Haonan He, Haodi Lei, Yun Luo et al.

AATC employs transform coding and reverse waterfilling for KV cache compression, achieving near-lossless accuracy at 5.8× compression.

Hannah Laus, Claudio Mayrink Verdun, Hao Wang et al.

P2Skill employs prompt-based skill distillation to enhance privacy preservation, achieving 1.69× higher inference quality without fine-tuning.

Myunghoon Ryu, Geunpyo Park, Sungjoon Lee et al.

HELIX employs model-harness co-evolution with source-traceable interventions, boosting task coverage by 4.0% and verified coverage by 58.0% in code repair tasks.

Tianyu Fan, Chao Huang

CForce enhances early-stage prediction reliability via self-trajectory consistency, significantly improving high-parallelism decoding in dLLMs.

Yuji Ren, Chenkai Xu, Zhuocheng Gong et al.

Longitudinal analysis of 33,228 PRs from vLLM and SGLang shows AI-assisted development boosts throughput, contributor diversity, and collaboration signals.

Jiada Li, Xuesong Ye, Olamide Olowoniyi

Event-driven Transformer-based DRL framework for multi-depot dynamic VRP, trained with behavior cloning and PPO, improves stability and responsiveness.

Faezeh Ardali, Gerald M. Knapp