MedFlow is a class‑aware multi‑scale flow matching framework designed to synthesize medical time‑series data. It uses a vector‑quantized multi‑scale tokenizer to capture both coarse and fine temporal patterns, and introduces Token Marginal Guidance to steer generation toward minority‑class characteristics. Experiments on four public datasets show MedFlow outperforms diffusion baselines, improving AUPRC by 5.8%, reducing Context‑FID by 88.6%, and achieving 3.8× higher sampling throughput.
By Yanhao Huang, Shibo Feng, Wanjin Feng, Peilin Zhao, Chunyan Miao
LookThere! Sparse Vision by Reinforced Selection proposes an end‑to‑end reinforcement learning framework that jointly trains a shallow input selector and a deep representation extractor for vision transformers. The selector learns where to focus and the extractor learns what to process, enabling the model to use only a tiny fraction of the input tokens—down to 0.2%—while preserving accuracy. The method outperforms existing selection techniques across diverse tasks and models, including high‑resolution recognition, segmentation, zero‑shot classification, and regression, establishing a new Pareto frontier in performance‑compute trade‑offs.
By Sreehari Rammohan, Yousef Yassin, Anthony Fuller, Junfeng Wen, Carl Vondrick, Evan Shelhamer
The paper introduces a framework for instruction‑guided 3D editing that does not require paired 3D supervision. It distills visual, semantic, and geometric knowledge from foundation models into a 3D editing model using a differentiable rendering pipeline, guided by a 2D visual prior from an image editing model and a semantic prior from a Vision‑Language Model. A 3D‑aware Distribution Matching regularization is added to prevent geometric collapse and ensure realistic 3D outputs, leading to superior instruction fidelity and cross‑view consistency compared to state‑of‑the‑art baselines.
By Hao Wen, Weibin Yun, Hongxing Fan, Haotian Lu, Rui Chen, Zehuan Huang, Lu Sheng
The paper presents an empirically calibrated, state‑aware Dynamic Voltage and Frequency Scaling (DVFS) scheduler that eliminates thermal throttling on a passively cooled Raspberry Pi 5 during sustained YOLOv8n inference. By using time‑domain guards, absolute temperature bounds, and derivative triggers, the scheduler outperforms a temperature‑only baseline with a 6.8% higher frame rate and 1.9% less energy per frame, and it surpasses an actively cooled reference in energy efficiency. The study also identifies that the passive operating envelope closes at ambient temperatures above 27 °C, where nonlinear leakage undermines DVFS control, and demonstrates that correct scheduling can make mechanical cooling unnecessary within the mapped envelope.
By Aayush Marasini, Zhaoxian Zhou
The paper documents a growing trend of removing safety guardrails from open-weight AI models, profiling the ecosystem that produces, redistributes, and applies these models. Between January 2024 and March 2026, 3,471 original uncensored models were identified on HuggingFace, each re‑packaged an average of 2.4 times, with three actors responsible for 52 % of all 8,164 compressed redistributions. After quantization and mirroring across platforms such as Ollama, these models persist even when upstream versions are removed, and 25 % of 1,643 GitHub applications that integrate uncensored large language models were classified as explicitly malicious.
By 10a Labs, :, Juliette Garcia, Hailey May, Bobby McKenzie, David Pham, Matthew Swain, Joshua Valdez, Corie Wieland, Zachary Yahn
The paper introduces Q‑MET, a quantum‑assisted framework that uses a hybrid quantum‑classical neural network to generate parameters for Wi‑Fi‑based human activity recognition models, drastically cutting the number of trainable parameters. By combining this approach with structured pruning during training, Q‑MET achieves 90–95% fewer trainable parameters than traditional backpropagation while preserving or improving classification accuracy. The method also yields 75–85% model sparsity for lightweight inference with less than 2% accuracy loss, enabling deployment on resource‑constrained devices.
By To Truong An, Jie Zhang, Guolin Yin, Junqing Zhang, Yanjiao Li, Trung Q. Duong, Simon L. Cotton
Hybrid language models combine attention with a fixed-size recurrent state, yet the distinct roles of each component are not well understood. The authors introduce two cache-level interventions—split-prefill and state-swap—to isolate the contributions of the KV cache (attention) and the recurrent state. Experiments on Qwen3.5 and Falcon-H1 show that exact retrieval depends almost entirely on attention, while output language and persona rely mainly on recurrence, with the state-swap intervention confirming that answers derive from the KV side and language from the recurrent side.
By Kirill Afendulev, Alexey Dontsov, Elena Tutubalina, Anton Korznikov
The paper introduces a free pause token that allows a language model to perform an extra parallel prediction stream without adding to the sequence length. This technique improves next‑token accuracy by 2–3 centi‑nats on a 1B‑parameter model while adding negligible inference latency or context overhead. Training costs increase modestly, with compute overhead reduced to about 1.14× compared to an optimized pretraining pipeline, preserving most benefits.
By John Langford, Nathan Godey, Giovanni Monea, Yoav Artzi, Harry Dong, Ying Fan, Gustavo de Rosa, Zheng Zhan
The paper examines how AI coding agents are evolving beyond simple autocomplete to perform complex tasks such as repository inspection, multi-file editing, tool execution, test writing, pull request creation, and long-duration work with minimal supervision. It highlights that while these agents boost coding activity, significant bottlenecks remain in review, integration, testing, security, deployment, and production operations, and that the economics of software development are shifting toward variable token, tool, sandbox, CI, and rework costs. The authors synthesize recent research and industry data to propose four engineering concepts—Agentic SDLC Throughput Paradox, Production-Qualified Change, Verification Tax, and an Agentic SDLC Control Plane—to guide the allocation of autonomy within cost, reliability, and human-attention constraints, ultimately reframing the research focus to production-qualified value per dollar, reviewer-hour, and operational risk.
By Happy Bhati
The paper introduces the concept of recurrent‑state write‑back, the rule that determines how a quantized state is stored and retrieved in recurrent neural networks. Using a compact GRU encoder–decoder for fluorescence lifetime imaging, the authors show that replacing continuous state propagation with deterministic 4‑bit storage dramatically increases estimation errors for lifetime parameters. They demonstrate that error feedback, residual memory, and direction memory can recover accuracy without retraining, and that the phenomenon also appears in LSTM networks, highlighting the importance of the state‑storage interface in low‑precision recurrent inference.
By Ismail Erbas, Xavier Intes, Vikas Pandey
The paper presents a unified compression framework for Vision Transformers aimed at on‑device plant disease detection in resource‑constrained agricultural settings. It combines Hessian‑Balanced Adaptive Block Pruning, quantization, and attention‑based knowledge distillation, evaluating each component separately before integrating the best performers into a deployment pipeline. On a chilli disease dataset, the compressed models achieve accuracy comparable to the FP32 baseline while reducing model size by 74‑98 %, and the full pipeline attains a 54.5× size reduction to 6.01 MB with 95.13 % accuracy.
By Mahadev Sunil Kumar, Bhavika Gondi, Desaisetty Venkata Satya Sai Swapnith, Gangireddy Rahul Jogi, Sudheesh Manalil, Arnab Raha, Amitava Mukherjee, Parthasarathy Seethapathy, G. Gopakumar
Weather-Conditioned Depth Anything (DA‑W) is a new framework that enhances monocular depth estimation models, like the Depth Anything series, to perform robustly under adverse weather conditions such as fog, rain, snow, and low‑light. It achieves this by disentangling style from content: a Style Filter extracts weather‑specific embeddings from a curated mix of real and synthetic degradation data, which are then injected into the backbone via a lightweight, zero‑initialized adapter. The adapter is trained with pseudo‑label distillation and alignment, enabling a single unified model to adapt to diverse weather scenarios while preserving its generalization on clean data, and it achieves state‑of‑the‑art performance with an average 3.7% improvement in AbsRel on weather benchmarks.
By Zhaoming Xu, Chan-Wei Hu, Kuan-Ru Huang, Zihao Zhu, Renjie Li, Yang Zhou, Zhengzhong Tu
Compact Bellman-Grounded Cognitive Maps (BCM) are introduced as a new method for cost-aware navigation that reuses a single learned map for different goals without per-goal retraining. BCM grounds the map in local edge costs using a self-supervised Bellman objective and a compact coordinate encoding, achieving near-optimal performance on weighted grids up to 1600 nodes with only a 5% mean gap to exact Dijkstra search. Its memory footprint grows sublinearly with graph size, outperforming connectivity-based spectral baselines and demonstrating scalability to complex environments.
By Yuzhe Han, Mingkun Xu, Yujie Wu
The paper introduces a method to compute Gaussian kernel sums—central to many kernel-based machine learning techniques—using flash attention, a highly optimized softmax attention implementation. By applying two small input augmentations, the authors transform the normalized softmax reduction into an unnormalized Gauss sum, eliminating the need for custom GPU code. For feature dimensions greater than eight in fp16, this approach outperforms both compiled PyTorch code and PyKeOps kernels in speed, memory overhead, and accuracy, while maintaining linear memory scaling.
By Nicolaj Rux, Sebastian Neumayer
CUA-Universe is a scalable environment-to-data pipeline that transforms real desktop software into hybrid GUI+CLI environments, enabling agents to coordinate visual inspection with command-line operations. It includes App-Forge for reproducible VMs and CLI surfaces, Task-Weave for generating diverse hybrid tasks, and Path-Steer for efficient rollouts and trajectory harvesting. Training on this data improves agent success and efficiency across multiple benchmarks, demonstrating the value of hybrid interaction.
By Haoting Shi, Wenhao Wang, Weicheng Fang, Yaozhong Liang, Tian Jin, Pengxiang Zhao, Guangyi Liu, Siheng Chen, Yanfeng Wang
ConsensusBench is a new dataset that supplies rule‑based process‑level signals for large language model reasoning. It identifies key intermediate conclusions—called Consensus Nodes—by filtering correct trajectories and clustering semantically equivalent statements. By incorporating a process reward derived from these nodes into GRPO‑style reinforcement learning, the authors create ConsensusPR, which reduces reward sparsity and improves performance on benchmarks such as AIME, GSM8K, and MATH‑500.
By Shi-Qi Yan, Chao-Hong Tan, Qian Chen, Wen Wang, Xiangang Li, Zhen-Hua Ling
The paper introduces AdaGate-DF, an adaptive gated deepfake detection framework designed for low-resolution and resource-constrained environments. By leveraging image-quality cues, the system routes samples through a dual multi-exit architecture, allowing high-quality images to exit earlier and reduce computational load. Evaluations on Celeb-DF and FaceForensics++ show that AdaGate-DF outperforms existing models such as MaD-CoRN and DefakeHop++ while maintaining low inference latency and robust performance across varying resolutions and class imbalance scenarios.
By Vaishnavi Sen, Cody Laurie, Rashida Hasan
The study explores non‑smooth activation functions—chaotic, stochastic, and fractal—in echo state networks, testing 36,610 reservoir configurations. It finds that functions like the Cantor function preserve the Echo State Property (ESP) even at spectral radii up to 10, outperforming traditional smooth activations in convergence speed. The authors introduce a theoretical framework for quantized activations, defining a Degenerate Echo State Property (d‑ESP) that implies the traditional ESP, and show that preprocessing topology, rather than continuity, governs stability.
By Rae Chipera, Jenny Du, Irene Tsapara
The paper presents a deep learning approach that converts nanopore resistive pulse signals into scaleograms using continuous wavelet transforms, enabling the classification of peptides as an image‑classification problem. On a dataset of 42 peptides, the method achieves an 82% macro‑averaged accuracy, outperforming previous descriptor‑based techniques by 8.6 percentage points. The models also remain accurate after significant weight pruning and 8‑bit quantization, making them suitable for deployment on embedded sensing hardware.
By Julian Ho{\ss}bach, Samuel Tovey, Sandro Kuppel, Tobias Ensslen, Jan C. Behrends, Christian Holm
Scale-QLoRA introduces a method for merging LoRA adapters into native 4‑bit microscaling language models without altering the quantized code plane. By training only the per‑block scale field and freezing the E2M1 code, the merged model remains code‑invariant and can be deployed as a single checkpoint. Experiments on four models and tasks show that Scale‑QLoRA achieves accuracy‑lossless merging comparable to merge‑aware QAT‑LoRA, while offering benefits such as exact rollback, code‑plane deduplication, and a 125× faster scale‑only task swap.
By Tung-Ling Li, Jiale Huang, Lee-Chi Wang, Janaki Ram Gotei