The study investigates how pruning affects large language models (LLMs) used for smart‑home tool calling. Researchers examined four LLMs—dense Transformer, dense hybrid, and mixture‑of‑experts (MoE) architectures—using depth, width, hybrid, and expert pruning, followed by supervised fine‑tuning. They evaluated over 19,500 instances from three smart‑home datasets, analyzing not only overall accuracy but also degradation in action components (operation, device, argument, value) and task complexity, finding that dense models suffer sharp performance drops after a narrow safe pruning range, while MoE models tolerate more pruning; aggressive pruning also leads to over‑refusal and loss of grounded specificity.
By Congjing Zhang, Vashishtha Patil, Henning Lange, Usman Aleem
arXiv:2609.16450v1 Announce Type: cross
Abstract: Diffusion large language models (dLLMs) offer a promising parallel decoding paradigm as an alternative to autoregressive generation through iterative...
By Lixuan Wei, Wei Zhou, Jianwen Wu, Yipeng Shen, Meiling Wang, Haoran You
OPD‑Aha is a privileged on‑policy distillation method that improves multimodal reasoning by reconstructing the distillation target from the teacher’s isolated visual preference instead of relying on fragile teacher‑student discrepancies. It suppresses continuations that contradict the image, encouraging students to interrupt flawed reasoning with reflection tokens such as "wait" and "actually." This approach leads to consistent improvements across fine‑grained perception and complex multimodal reasoning benchmarks.
By Chenhao Qiu, Dawei Li, Yechao Zhang, Lei Gong, Zhen Tan
arXiv:2609.16255v1 Announce Type: cross
Abstract: We present an efficient method to distill reasoning capabilities into compact video-language models (VLMs) for video question answering (VideoQA). Ou...
By Mantek Singh, Jeshwanth Challagundla, Siddharth Raina, Jasmin Jarsania
arXiv:2609.17193v1 Announce Type: new
Abstract: Large language model (LLM)-powered agentic AI services increasingly demand low-latency inference, motivating the deployment of LLMs across distributed...
By Zhen Li, Jun Cai, Haoran Gao, An Li, Tan Li
arXiv:2609.17019v1 Announce Type: new
Abstract: While Chain-of-Thought (CoT) reasoning has been proven to be effective, it often leads to overthinking, resulting in computational overhead, inference...
By Qinhong Lin, Yuhao Zhang, Yinglun Feng, Zhongliang Yang, Linna Zhou
arXiv:2609.16215v1 Announce Type: new
Abstract: GPU high bandwidth memory is scarce and expensive, and KV caches consume much of it as chats, agent loops, and document question answering accumulate s...
By Srikanta Datta Tumkur, Jay Iyer, Mehar Simhadri, Sai Pavan Kumar, Sai Kapil Kumar, Ramesh Nampelly
The paper introduces LCAP, a method for adapting photonic neural networks to real hardware by learning a shared correction from a population of chips and then personalizing each chip using only 32 fixed output probes. LCAP decomposes adaptation into a transferable population correction and a probe‑inferred latent personalization, allowing feed‑forward calibration without device‑specific optimization. Experiments on a simulated three‑layer 64‑mode MZI network show accuracy improvements from 80.4% to 93.4% and significant gains on unseen chips.
By Tianyu Gao, Guantian Zheng
arXiv:2609.17391v1 Announce Type: new
Abstract: Model serving is one of the largest cost drivers in production recommender systems. Maximizing its throughput requires navigating a deeply layered hier...
By Qi Wu, Lohan Lemire, Kai Meng, Zhongmou Cai, Raphael Bargues, Petr Zhitnikov, Zeyuan Cao, Yao Wang, Shujun Bian, Wei Chen, Sean Sheng
The paper introduces MedPCFM‑TED, a one‑step distillation framework that uses teacher‑guided endpoint supervision and geometric matching losses to generate cranial implants from point clouds. It outperforms existing one‑step methods on the SkullBreak benchmark, remains competitive on SkullFix, and achieves a generation time of about 0.04 s per sample. The approach demonstrates that rapid, high‑quality implant generation is possible without multiple neural evaluations during inference.
By Kamil Kwarciak, Marek Wodzinski
arXiv:2609.17109v1 Announce Type: new
Abstract: A common small-model deployment runs one shared backbone with several LoRA specialists that answer over the same context. Serving them naively re-prefi...
By Dushyant Rajput
The article surveys how large language models (LLMs) can be incorporated into networked control systems, cyber‑physical systems, and multi‑agent networks without violating stability and safety guarantees. It proposes treating the LLM as a slow supervisor that sets high‑level goals, while a fast, certified inner loop preserves physical stability. The survey maps LLM characteristics—such as inference latency, API failures, tokenization, and hallucinations—to classical control challenges and highlights the growing gap between model capability and formal safety assurances, calling for future research on stability proofs.
By Haiping Du, Linping Chan
The paper re‑examines the impact of removing the encoder from Action Chunking Transformers (ACT), a model used for robot manipulation learning. Contrary to the original claim that encoder removal drops success rates from 35% to 2%, the authors find no such dramatic effect in their re‑runs, though minor variations remain uncertain. They attribute discrepancies to training length and checkpoint selection, and note that the encoder’s latent variable offers little reconstruction benefit on the tested benchmark, while its removal speeds up training.
By Bo Kang
The paper introduces Adaptive Derivative-Ordered Random Explanation (ADORE), a unified framework that uses first- and second-order derivatives to capture nonlinear feature interactions and feature-sample dynamics. ADORE combines global feature importance with local sample contributions, quantifying both magnitude and direction of feature impact while identifying critical samples. It achieves computational efficiency via randomized SVD and dynamic sparsity detection, outperforming LIME and SHAP across tabular, text, and image data, and is released as an open-source Python package on GitHub.
By Lemen Chao, Ming Lei, Anran Fanga
arXiv:2609.17509v1 Announce Type: cross
Abstract: Neural audio codecs are a key component in speech language modeling. However, their high frame rates lead to long sequence lengths, increasing comput...
By Thanapat Trachu, Samuele Cornell, William Chen, Shinji Watanabe
arXiv:2609.17026v1 Announce Type: new
Abstract: Continual learning must balance the learning of new knowledge with the retention of previously learned knowledge to incrementally learn tasks from a da...
By Yunxiang Fu, Meng Lou, Zicheng Liao, Yizhou Yu
The paper evaluates a learned request‑routing policy for disaggregated large‑language‑model serving, where compute‑heavy prefill and memory‑heavy decode stages run on separate GPU pools. Using a discrete‑event simulator and real NVIDIA A40 GPUs, the calibrated router—leveraging prompt length, predicted output length, KV‑cache pressure, and SLO class—outperforms round‑robin, least‑loaded, and length‑based heuristics, achieving the highest mean goodput (0.864) and lowest variance across three mixed, bursty arrival traces. Hardware calibration proves critical, providing a 4.5‑point goodput boost and roughly 40 % of the tail‑latency advantage, and the learned router can match round‑robin performance with one fewer GPU in certain scenarios.
By Srikanta Datta Tumkur, Jay Iyer, Mehar Simhadri, Sai Pavan Kumar, Sai Kapil Kumar, Ramesh Nampelly
Large language model optimization is an active research area, spanning quantization of model weights, early-exit methods for skipping layers, and speculative decoding. Each track uses its own quality...
We study the problem of demonstration selection, which involves selecting a subset of examples for prepending to a query to a language model. This problem is closely related to in-context learning and...
Neural audio codecs are a key component in speech language modeling. However, their high frame rates lead to long sequence lengths, increasing computational costs. Dynamic frame rate codecs mitigate t...