REALM is a retrospective knowledge distillation framework that enables causal decoding of behavior from local field potentials (LFPs). It trains a bidirectional Mamba‑2 teacher on multi‑session data using continuous masked autoencoding, then distills its representations into a compact causal student model. The resulting LFP‑only decoder achieves the highest mean accuracy among compared methods, surpassing state‑of‑the‑art baselines while using fewer parameters and less pretraining time.
By Peicheng Wu, Zhenyu Bu, Runze Ma, Lin Du
GPart introduces a new parameter‑efficient fine‑tuning technique that directly maps a low‑dimensional trainable vector into the full weight space using a sparse, isometric partition matrix. Unlike LoRA, GPart eliminates the bilinear reconstruction step, preserving exact end‑to‑end isometry and reducing the checkpoint to just the vector and a random seed. Experiments across NLP, vision, and reasoning tasks show that GPart matches or surpasses existing PEFT methods while using far fewer parameters and offering a simpler, more tractable parameterization.
By Paolo Mandica, Micha{\l} Brzozowski, Zuzanna Dubanowska, Neo Christopher Chung
The paper introduces GAP-DPO, a method for personalizing large language models by selecting preference pairs based on gradient alignment with user utility. It formalizes personalized preference learning as a geometry‑aligned optimization problem, showing that off‑policy sampling can shift DPO updates from error correction to reinforcement when preference margins align with utility gradients. Experiments demonstrate that GAP‑DPO improves stylistic fidelity, preference alignment, and overall generation quality over standard DPO variants.
By Ruoming Jin, Xinyu Li, Hao Zhou, Jianfeng Zhu, Ruixin Guo, Feodor Dragan, Lei Xu, Haixun Wang, Yang Zhou
EviGraph is a system that enables proof‑carrying selective recommendation over temporal public‑service knowledge graphs. It distinguishes between critical decision requirements and information that can remain unresolved, linking the former to evidence via a language agent and verifying support with a deterministic checker. Experiments on a bilingual Hong Kong public‑service benchmark show that this approach reduces unnecessary abstention, while additional verification can actually withdraw supported recommendations without improving decision quality.
By Yixi Zhou, Sikun Wang, Lei Fan, Fan Zhang
arXiv:2610.02181v1 Announce Type: new
Abstract: We present OmniSeek, an agentic framework that transforms an Omni Large Language Model (Omni-LLM) into an active, multi-turn reasoning agent with nativ...
By Haibo Wang, Jiteng Mu, Jialu Li, Jingru Yi, Yuanjun Xiong, Jianming Zhang, Lifu Huang, Mingze Xu
LensVLM is an inference framework and post‑training recipe that lets Vision‑Language Models (VLMs) process compressed images of text by selectively expanding only the relevant parts back to full resolution. Using Qwen3.5‑9B‑Base, LensVLM achieves accuracy comparable to full‑text models at 4.3× compression and outperforms other compression baselines up to 10.1× across seven text QA benchmarks, while also improving performance on multimodal document and code tasks as compression increases.
By Roy Xie, Dan Friedman, Donghan Yu, Bowen Pan, Christopher Fifty, Jang-Hyun Kim, Xianzhi Du, Zhe Gan, Vivek Rathod, Bhuwan Dhingra
The study evaluates how scientific explanations influence experimental forecasts by comparing paired predictions that share the same intervention, forecaster, and outcome but differ in description, matched explanation, and donor context. Across 336 prospective states, 12 Tox21 endpoints, and 24 OpenML tasks, the predictive credit of explanations was largely inconclusive, with key metrics such as ROC AUC interval-score and point-equivalence rules unmet. While certain model variants (e.g., DeepSeek V4 Pro) reduced drift and improved some error metrics, overall matched point-accuracy gains and natural-explanation credit remained unconfirmed at the tested donor resolutions.
By Jingjie Ning, Xueqi Li, Yibo Kong, Dongting Li
VeriHarness is a method that enhances verification for large language model agents tackling long‑horizon tasks without needing reference answers at test time. It transforms the base LLM into an agentic verifier by providing a workspace, evidence tools, and reusable verification skills, using disagreement resolution and consensus challenge to evaluate competing claims. Across five benchmarks and two frontier models, VeriHarness outperforms baselines, achieving significant performance gains and demonstrating self‑improvement of verification skills from failure feedback.
By Caiqi Zhang, Rujun Han, Zifeng Wang, Zoey CuiZhu, Nigel Collier, Tomas Pfister, Chen-Yu Lee
Ireland’s smart metering programme records electricity use at 30‑minute intervals, which is too coarse to capture domestic appliance use. The authors present a label‑free disaggregation system that splits usage data into nine appliance categories by combining event detection for high‑power loads with questionnaire‑guided estimation. Evaluated on four datasets—including a large Irish smart‑meter dataset of over 4,800 years of use—the hybrid method achieves the lowest whole‑decomposition error and better month‑level performance than two independently developed systems.
By Achal Nanjundamurthy, Rupam Misra, Suzanne Little, Alan F. Smeaton
VideoEvolve is a framework that automatically evolves agent harnesses for video temporal grounding, a task that localizes events in videos based on natural-language queries. It introduces a Cloze-Structured Harness Representation to keep stage interfaces stable while allowing agent workflows and instructions to change, and uses Branch-Guided Harness Evolution to preserve promising code branches, guide local edits with execution feedback, and validate improvements. Experiments show that this automated evolution improves grounding performance across multiple benchmarks, with instruction refinement consistently yielding gains.
By Bingjun Luo, Yuhuan Fan, Jialin Guo, Siqi Li
The paper introduces Quasar, a model‑free Q‑learning algorithm that guarantees asymptotic convergence for reachability objectives in Markov Decision Processes that are free of non‑terminal maximal end components (MECs). Unlike prior model‑based methods, Quasar does not estimate transition probabilities, reducing memory usage from O(|S|²|A|) to O(|S||A|). Experiments on the Quantitative Verification Benchmark Set show that Quasar converges to optimal policies with far fewer samples than existing state‑of‑the‑art model‑based approaches.
By Lu-Chin Chang, Suguman Bansal
The paper introduces the Homomorphic Advantage Operator (HAO), a stabilization framework for reinforcement learning under fully homomorphic encryption (FHE). HAO applies a zero‑mean centering projection to temporal‑difference targets, eliminating the uniform state‑value baseline that causes the Bellman drift and keeping network pre‑activations within safe polynomial approximation bounds. Experiments on tabular MDPs, an encrypted CartPole environment, and a logistics routing benchmark show that HAO prevents boundary breaches entirely, improves optimal policy accuracy by 18 percentage points, and remains stable even with added Gaussian noise.
By Abid Mohamed Nadhir, Ahmad Al Hanbali, Beggas Mounir
Tokenized Key-Gated Adapter Routing (Locket) is a framework that embeds fine‑grained, policy‑driven access control into large language models by training lightweight LoRA adapters for different privacy policies. A gating module associates a learned keyed entry token with a specific adapter, allowing authorized tokens to unlock private knowledge while invalid or missing tokens trigger privacy‑preserving adapters that redact or sanitize sensitive content. Experiments on datasets such as Enron, ECHR, and Yelp with models like Qwen3, Llama‑3.2, and Gemma‑2‑2B show that Locket maintains perplexity comparable to fine‑tuning when the correct token is provided, and significantly reduces PII leakage when the token is absent or invalid, without sacrificing utility.
By Mohamed Shaaban, Mohamed Elmahallawy
DuplexSpeechBench-Document Grounding (DSB‑DG) is a benchmark that evaluates how well voice agents can ground their responses in external documents across five professional domains. It focuses on three failure modes—Context Saturation, Grounding Decay, and Proactive Grounding—and includes 1,636 adversarially verified QA pairs from 50 documents. The benchmark measures grounding accuracy, hallucination, and response latency, revealing that cascaded pipelines achieve the highest accuracy while open‑weight systems suffer from abrupt capacity collapse and multi‑turn decay.
By Puneet Mathur, Nedim Lipka, Zeyu Jin, Dinesh Manocha
The paper investigates how the set of prior iterates influences success in large language model (LLM)-driven discovery tasks. It introduces 12 new harnesses called 'Modular' and evaluates them on five diverse discovery problems, revealing that success is fragile and highly dependent on harness design. The study identifies mode collapse—a sharp loss of iterate diversity—as a common failure, and shows that early discoveries predict final outcomes, leading to a new initialization strategy that consistently improves performance across harnesses and applications.
By Mansi Sakarvadia, Marco Ciccone, Colin Raffel
This study introduces the first controlled benchmark of generative models for weather data assimilation using real station observations from 11,849 NOAA MADIS stations across the U.S. It evaluates key design choices—diffusion vs. flow matching, pixel vs. latent-space formulations, and inference-time conditioning strategies—against a classical 3D-Var baseline. The benchmark finds that learned generative priors and full-gradient guidance improve RMSE over ERA5, while other design variations offer minimal benefit, especially under sparse observation conditions.
By Ruizhe Huang, Qidong Yang, Jonathan Giezendanner, Sherrie Wang
The paper reports that applying reinforcement learning (RL) to the Kimi K2.7 Code model on 1,700 agentic coding tasks improves its performance on six external benchmarks. After a single epoch of GSPO training on a rank‑32 LoRA adapter, pass‑@1 scores increased across all benchmarks, with significant gains even on data released after training. The trained model also reduces agent steps and avoids common failure modes such as dropping requirements or breaking existing behavior.
By Sushant Mehta, Logan Ritchie, Edwin Chen
MWOP (Modality-aware Width-wise Operation Pruning) is a method that independently prunes visual‑to‑visual, text‑to‑visual, and text‑to‑text attention paths within each layer of multimodal large language models, and separately selects feed‑forward network channels for visual and textual inputs. It uses a first‑order Taylor criterion to guide pruning, re‑evaluates FFN importance after attention pruning, and applies LoRA‑based recovery training. The approach is paired with path‑sparse Triton attention kernels and compact visual‑side FFN execution to achieve practical acceleration, preserving token sequences while reducing computation.
"whyItMatters":"MWOP achieves a 1.6× prefill speedup on LLaVA‑OneVision‑7B while retaining 99.7% performance, and further boosts token‑compression methods to 2.9× and 2.7× speedups, demonstrating its effectiveness across architectures."
By Xudong Wang, Hao Wu, Haozhe Hu, Peiran Yin, Xinghao Chen, Yunpu Ma, Wei Zhang, Xiaoyu Shen
The paper introduces a multi‑party backchannel prediction benchmark built from the AMI meeting corpus, featuring 682 masked‑listener views, 190 speakers, and 18,697 backchannel events. A state‑of‑the‑art dyadic model performs at chance when applied zero‑shot to meetings, but its frozen acoustic features are still informative, and retraining improves performance to an AUROC of 0.751. The study reveals that listener conditioning helps only for listeners seen during training, that speaker identity is entangled with useful cues, and that backchannel rates vary significantly across individuals, prompting the authors to report both AUROC and event‑F1 metrics.
whyItMatters":"The benchmark and evaluation tools provide a standardized, person‑disjoint testbed for advancing multi‑party backchannel prediction research."
By Mohammed Hafsati, Ahmed Loughzali
The paper examines safety routers—systems that route user requests to different language models—and finds that their performance degrades significantly when evaluated under distribution shift. In standard benchmarks, routers appear effective because the best single model is chosen from the same evaluation data, but when the data distribution changes, the routing advantage diminishes or disappears. The study quantifies this bias across multiple safety corpora, showing that routers offer little benefit under realistic shift conditions and that recognition‑based defenses can be undermined by attackers who know the model being used.
By Amit Singh Bhatti, Vishal Vaddina