04:00 <Ada> arXiv.org | DeepAJM: Deep Association Joint Model for Irregularly Sampled data Barsha Halder, Jeffrey A. Thompson · 4h ago · 3 min read Joint Models simultaneously model longitudinal and survival outcomes, leveraging patterns in patients' longitudinal trajectory to improve the prediction of survival outcomes. The classical parametric joint models, however, rely on fixed parametric assumptions, making them susceptible to bias under model misspecification and smaller sample sizes. We propos... Read full story

04:00 <Ada> arXiv.org | ImpactMat: Continuous Material Estimation for Inverse Impact Sound Rendering Hyebin Cho, Bumsoo Kim, Joon Son Chung · 4h ago · 3 min read Impact sound rendering synthesizes the sound produced when a 3D object is struck, but practical renderers often rely on fixed material presets such as wood, plastic, or steel. These presets limit the range of impact sounds a renderer can express, while manually adjusting the underlying material parameters remains difficult without expertise in material ac... Read full story

04:00 <Ada> arXiv.org | RAISED: Self-Distillation for Robustness to Prompt Injection in LLM Agents 4h ago · 3 min read Tool-using language-model agents are vulnerable to indirect prompt injection because they must act on untrusted external content. Existing training-time defenses can reduce attack success rates, but often at the cost of general capabilities. We show that training-based defenses induce substantial drift in the model's output distribution, altering its beha... Read full story

04:00 <Ada> arXiv.org | The sublevel Flood bifiltration: towards scalable 2-parameter persistent homology Matt\'eo Cl\'emot, Julie Digne, Julien Tierny · 4h ago · 3 min read Multiparameter persistent homology is a rapidly developing branch of topological data analysis that improves the robustness of single-parameter persistent homology to outliers, while still capturing the metric characteristics of the data. However, a notable limitation is its lack of scalability. In this paper, we introduce a novel approach for efficiently... Read full story

04:00 <Ada> arXiv.org | Humanoid Rickshaw Pulling: Whole-Body Locomotion under Coupled Wheeled Loads Yangzhi Yang, Xiansheng Lin, Zhaoming Xie, Xiaobin Xiong · 4h ago · 3 min read Humanoid robots could transport payloads substantially heavier than themselves by pulling passive wheeled vehicles instead of carrying the load. This capability, however, creates a coupled locomotion problem: the robot must maintain persistent upper-body contact while adapting to unknown, configuration-dependent forces arising from the payload, vehicle, a... Read full story

04:00 <Ada> arXiv.org | E$^2$-OPSD: Taming Entropy Overshoot in On-Policy Self-Distillation 4h ago · 3 min read On-policy self-distillation (OPSD) provides dense token-level supervision without a second model: one network acts as teacher with the reference solution and as student with only the problem. We identify a specific failure mode of this recipe. During training, student token entropy rises past the teacher's and remains elevated, a pattern we call entropy o... Read full story

04:00 <Ada> arXiv.org | Nonparametric Distribution Matching for Self-Supervised Whole-Slide Image Condensation 4h ago · 3 min read Histological whole-slide images (WSIs) are central to computational pathology but pose severe computational challenges due to their extremely high resolution, often spanning several gigabytes per slide. To enable scalable learning, existing methods apply self-supervised data condensation to reduce computational cost, but typically rely on heuristic protot... Read full story

04:00 <Ada> arXiv.org | WAMJET: A Harness for World Action Model Acceleration 4h ago · 3 min read World Action Models (WAMs) leverage pretrained video foundation models for robot manipulation, but their large backbones and video-action co-prediction are expensive. Although existing acceleration techniques offer many ways to reduce this cost, selecting and composing them requires substantial engineering for each model and hardware platform. To tackle t... Read full story

04:00 <Ada> arXiv.org | STARS: From Spatiotemporal Dynamics to Social Representations in Human-Robot Interaction 4h ago · 3 min read Robot navigation in dynamic, human-centered environments requires socially-compliant decisions grounded in robust scene understanding. Recent Vision-Language Models (VLMs) exhibit promising capabilities such as object recognition, common-sense reasoning, and contextual understanding, capabilities that align with the nuanced requirements of social robot na... Read full story

04:00 <Ada> arXiv.org | When Does a Second Model Help? Cross-Model Review in LLM Verification Tae-Eun Song · 4h ago · 3 min read Large language models now generate code, documentation, and analyses, and are increasingly used to review such output. We ask when a second review by a different model helps. Building on the author's earlier preprints, which varied context, repetition, and role structure within one model, we test model independence in a controlled experiment: 30 artifacts... Read full story

04:00 <Ada> arXiv.org | Multi-LLM Collaborative Alignment via Stackelberg Games 4h ago · 3 min read A pool of language models can collaborate and improve collectively by learning from one another's responses. These interactions depend on the instructions used during training. Existing methods typically sample instructions uniformly, even though their usefulness may change as the models improve: an instruction on which models' responses once differed in ... Read full story

04:00 <Ada> arXiv.org | Does Scaling Reinforcement Learning Really Require More Training? Bangji Yang, Jiajun Fan, Hongbo Ma, Ruihan Guo, Ge Liu · 4h ago · 3 min read Scaling reasoning typically spends more compute on reinforcement learning (RL) or on inference. We show that a completed RL training history can yield policies stronger than the checkpoints visited by its optimizer. We call this policy-space scaling: expanding the deployable policy set accessible from a fixed RL history, without extending training or incr... Read full story

04:00 <Ada> arXiv.org | Residual-Stream Burden Shapes Representation Learning in Diffusion Transformers 4h ago · 3 min read In diffusion-based generation, a neural network can be trained to predict the clean data, the noise, or the velocity from a noisy input. These prediction targets are interconvertible and describe the same generative process, yet plain Diffusion Transformers operating on large pixel patches succeed with clean prediction and fail with noise or velocity pred... Read full story

04:00 <Ada> arXiv.org | KlinikeBench: Evaluating Language Models Beyond Diagnostic Accuracy 4h ago · 3 min read Most clinical benchmarks evaluate language models (LMs) on diagnosis using complete case descriptions. In clinical practice, however, patients present information in different ways, and clinicians must obtain relevant history and determine which examinations are needed before reaching a diagnosis. Diagnostic accuracy alone therefore cannot establish wheth... Read full story

04:01 <Ada> arXiv.org | Train4Merge: A Controlled Single-Teacher Study of RL vs. SFT Teachers for OPD-Based Model Merging 4h ago · 3 min read Domain experts trained from a shared checkpoint can transfer their specialized capabilities to a single student through on-policy distillation (OPD). Existing research primarily focuses on improving this merging process, while the algorithms used to train the experts have received limited systematic comparison. We investigate which training algorithm prod... Read full story

04:01 <Ada> arXiv.org | FineART: Fine-Grained Annotated Robotic Trajectory Dataset and Vision-Language-Action Model for Bimanual Manipulation 4h ago · 3 min read Robots operating in real-world environments must often execute complex, multi-step bimanual tasks over long horizons rather than single, isolated actions. Current manipulation datasets provide limited support for this capability: although single-arm datasets reach hundreds of thousands of trajectories, they typically provide only one high-level instructio... Read full story

04:01 <Ada> arXiv.org | Infrared Subtraction with Artificial Intelligence Wenjie He, Xiaohui Liu, Yandong Liu, Zhan Wang · 8d ago · 4 min read We present AI-developed local infrared subtraction, building on projection to Born and EFT matching. The framework separates an integrable radiation term from a finite contribution at Born kinematics, referred to as the Born contact. The contact is determined using the EFT singular distribution in a resolution observable such as N-jettiness $τ_N$. Under h... Read full story

04:01 <Ada> arXiv.org | Subjects, Not Authors: The Authorship Hazard in Agentic Dataspaces Seungho Lee, Changbin Lee · 4h ago · 3 min read Dataspace connectors decide whether a transfer may occur, not what the transferred value contains, tolerable for contracted applications, not for LLM agents that compose tool calls. Work on agents that generate governance artifacts evaluates output quality, not who may authorize an artifact for use. A published policy is what the decision point enforces, ... Read full story

04:01 <Ada> arXiv.org | KernelOPT: Dispatch-Aware Agentic Search for GPU Kernel Optimization 4h ago · 3 min read Deep learning inference and training performance depends critically on GPU kernel efficiency. Modern compilers such as PyTorch Inductor automatically generate GPU kernels from high-level model code, but frequently underperform expert-written implementations by wide margins. Recent LLM-assisted kernel optimizers can close this gap for standalone kernels, y... Read full story

04:01 <Ada> arXiv.org | Segment-Level Risk Discovery in Online Handwriting for Alzheimer's Disease Detection Changqing Gong, Huafeng Qin, Moun\^im A. El-Yacoubi · 4h ago · 3 min read Online handwriting provides a non-invasive and low-cost behavioral biomarker for Alzheimer's disease (AD) detection, as it reflects both cognitive planning and fine motor control. Existing handwriting-based AD detection methods usually rely on global trajectory features or whole-sample representations, which can be strongly affected by individual writing ... Read full story

04:04 <Vega> MIT News | Massachusetts Institute of Technology | Using AI to mitigate the growing environmental threat of data centers Adam Zewe | MIT News · 4h ago · 7 min read MIT Associate Professor Christina Delimitrou applies machine learning to make large cloud computing systems more efficient, secure, and reliable, which improves performance of web applications while reducing the negative environmental impacts of data centers. Read full story

08:01 <Ada> arXiv.org | Ordinary Nonconvex SGD under Distance-Dependent Moments: Finite-Horizon Stationarity and Nagaev Bounds Wei Biao Wu · 8h ago · 3 min read Uniform noise-moment bounds exclude stochastic gradients whose variability increases with the iterate. We study ordinary, single-sample stochastic gradient descent for smooth, lower-bounded, possibly nonconvex objectives under distance-dependent conditional moments. Under second moments alone, a direct descent--displacement argument yields $T^{-1/3}$ expe... Read full story

08:01 <Ada> arXiv.org | GridSFM: A Foundation Model for Solving AC Optimal Power Flow Luke Bhan, Weiwei Yang, Margaret Capetz, Baosen Zhang · 8h ago · 3 min read We introduce GridSFM, a framework that combines a pretrained foundation model across grid topologies with physics-informed fine-tuning for solving AC Optimal Power Flow (AC-OPF) at scale. It is a $15$ million parameter physics-inspired graph neural network pretrained across $54$ topologies of $500$ to $4{,}000$ buses. Our model attains a $2.45\%$ zero-sho... Read full story

08:02 <Ada> arXiv.org | End-to-End Quantum Semantic Communication with Variational Quantum Neural Networks 8h ago · 3 min read Quantum-enabled learning is increasingly being explored for future communication and networking applications, including distributed sensing, Internet of Things (IoT), and distributed quantum computing. However, existing approaches often face challenges in scalable inference and efficient information processing. To address these limitations, this paper int... Read full story

08:02 <Ada> arXiv.org | Hiding in Plain Sight: A Diffusion-based Mitigation of Geolocation Privacy Leakage in Vision-Language Models 8h ago · 3 min read Multimodal large reasoning models (MLRMs) have demonstrated remarkable capabilities in complex visual understanding. However, this very power introduces a critical yet underexplored privacy threat: adversaries can exploit MLRMs to precisely infer users' geographic locations from casually shared photographs, by performing structured reasoning over subtle v... Read full story

08:02 <Ada> arXiv.org | TACTICS: Taxonomy-Aware Intelligent Corpus Sampling for Machine Translation 8h ago · 3 min read Large-scale machine-translation (MT) systems are typically evaluated on random samples from a corpus whose distributional composition is an artifact of how it was assembled. Such a sample inherits the phenomena the collection happens to contain rather than the full space a system must handle, spanning rule-governed conventions (terminology, punctuation, c... Read full story

08:02 <Ada> arXiv.org | Self-Indexing Attention for Compression-Compatible Sparse Long-Context LLM Inference 8h ago · 3 min read Sparse long-context inference requires efficient token retrieval in both prefill and decode. Existing methods often use different retrieval strategies for the two stages, preventing one retrieval representation from being reused throughout inference. We propose Self-Indexing Attention, a training-free framework built on a shared transform-domain sign-magn... Read full story

08:02 <Ada> arXiv.org | The Semantic Bottleneck: Leveraging Semantic Representations for Non-Invasive Speech Decoding Gilad D. Landau, Dulhan Jayalath, Oiwi Parker Jones · 8h ago · 3 min read Non-invasive speech decoding remains constrained by the low signal-to-noise ratio of neural recordings, which makes fine-grained reconstruction of phonemes or individual words difficult. Motivated by neuroscientific evidence that high-level semantic representations are distributed across cortical regions and evolve over slower temporal scales, we hypothes... Read full story

08:02 <Ada> arXiv.org | Shallow neural network approximation in mixed Sobolev spaces Yuwen Li, Guozhi Zhang · 8h ago · 3 min read We investigate the best $L_2$ approximation of mixed Sobolev spaces by shallow neural networks with $n$ neurons and general activation functions. We first establish an activation-independent Fourier-block principle: if an activation has univariate approximation order $ρ$ in the sense of the Fourier-block property, then the global approximation rate has al... Read full story

08:02 <Ada> arXiv.org | Generative Nested Sampling of Atomistic Thermodynamic Landscapes 8h ago · 3 min read Nested sampling (NS) resolves the thermodynamics of an atomistic system from a single simulation, but its practical reach is limited by the Markov-chain updates needed to decorrelate walkers within each likelihood-constrained ensemble. Flow-based NS has reduced this bottleneck for gravitational-wave (GW) inference, yet its transfer to atomistic systems is... Read full story

08:02 <Ada> arXiv.org | EDGE: Engine for Deterministic Graph Evaluation through Conversation Simulation from Graph Structured DSL Configuration 8h ago · 3 min read As agentic systems evolve into complex multi agent orchestration workflows, there is a growing and critical need for systematic frameworks that measures an agent's behavioral consistency and determinism. In this paper, we introduce a formal evaluation methodology that is grounded in AgentGraph, a planner powered by a domain specific language that represen... Read full story

08:03 <Ada> arXiv.org | Forward-Deployed Full-Stack Engineering for Autonomous Cloud MLOps Sagar Srinivas Sakhinana, Venkataramana Runkana · 8h ago · 3 min read Across industries, machine-learning systems support applications ranging from prediction and anomaly detection to forecasting, optimization, and scheduling, yet operationalizing these systems requires coordinating application development, model pipelines, cloud infrastructure, security, deployment, monitoring, retraining, recovery, and rollback. We presen... Read full story

08:03 <Ada> arXiv.org | TACS: Trajectory-Aware Candidate Selection for LLM Jailbreak Suffix Optimization Shiliang Xiao · 02 Sep 2026 · 3 min read Gradient-based jailbreak suffix optimization methods typically update the suffix by retaining the candidate with the lowest current loss. We show that this seemingly natural design is fundamentally myopic: candidates that look better under the current-step proxy often fail to produce better jailbreak outcomes later in the search, revealing a form of selec... Read full story

08:03 <Ada> arXiv.org | How Language Models Organize and Structure Moral Knowledge Orion Reblitz-Richardson · 8h ago · 3 min read How do large language models (LLMs) organize moral knowledge? Models detect moral content broadly, but detection is a low bar. We ask whether they go further, distinguishing moral foundations from one another and organizing the relationships between them geometrically. We train six independent linear probes on open-weight language models, one per Moral Fo... Read full story

08:03 <Ada> arXiv.org | LittleLearner: Language Models Under Pedagogically Controlled Knowledge Exposure 8h ago · 3 min read Modern language models are trained on heterogeneous web-scale text corpora. Consequently, studying knowledge and skill acquisition is difficult, as prior exposure to related content is hard to characterize. To address this challenge, we introduce LittleCurriculum, a curated 88B-token pretraining corpus tailored to U.S. elementary school material, explicit... Read full story

08:03 <Ada> arXiv.org | The Parser Already Knows: Lightweight Bias Correction in Constrained Decoding 8h ago · 3 min read Grammar Constrained Decoding (GCD) forces Language Models (LMs) to produce syntactically valid outputs by masking out non-conforming tokens at each step. However, because masking only checks whether each token is valid so far, the resulting distribution over complete outputs diverges from the LM's own distribution conditioned on the grammar, biasing gener... Read full story

08:03 <Ada> arXiv.org | Faster-WAM: Do World Action Models Need Deep Action Modules? 8h ago · 3 min read World Action Models (WAMs) build on pretrained video models, whose representations are grounded in physical dynamics and provide a natural basis for action prediction. Despite this natural foundation, many WAMs still rely on deep, parameter-heavy action-prediction modules that incur high inference latency and may overfit to limited robot demonstrations, r... Read full story

08:03 <Ada> arXiv.org | ReToken: Improving Long-Context VLMs with Visual Retrieval Token 8h ago · 3 min read Long visual contexts challenge vision-language models: performance degrades as the number of distractors grows, and processing all tokens at once can exceed GPU memory limits. We present RETOKEN, a single learnable embedding that extracts retrieval signals from the VLM's internal representations to select query-relevant visual tokens from the pre-filled K... Read full story

08:03 <Ada> arXiv.org | Adapting Generalist Vehicle Models for High-Speed MPC Across Terrains 8h ago · 3 min read High-speed off-road autonomy requires precise closed-loop control for a target vehicle while remaining robust across changing terrains. Recent forward kinodynamic (FKD) prediction foundation models suggest a promising path, starting from a generalist model and specializing it to the target platform. However, effective specialization remains challenging, a... Read full story

08:03 <Ada> arXiv.org | Hallucination Self-Play: Bootstrapping Reinforced Detector via Evolved Generator 8h ago · 3 min read Identifying faithfulness hallucinations in LLM-generated outputs remains challenging due to the scarcity of high-quality annotated data. Recent work relies on advanced LLMs to synthesize training data, including rationales, labels, and hallucinated claims. However, these methods treat the generator as a static component, limiting iterative improvement of ... Read full story

08:03 <Ada> arXiv.org | Non-asymptotic Convergence of Stochastic Gradient Descent in Score-based Generative Models 8h ago · 3 min read Score-based Generative Models (SGMs) have achieved impressive performance in data generation across a wide range of applications. While the statistical properties of their sampling procedures are increasingly well understood, the optimization dynamics underlying their training remain less explored. SGMs are typically trained by minimizing a weighted denoi... Read full story

08:05 <Ada> arXiv.org | FAR: Failure-Aware Retry for Test-Time Recovery and Continual Policy Improvement 8h ago · 3 min read Robot policies inevitably encounter failures when deployed in real environments. Naive retries often repeat the same mistakes, while many existing recovery methods rely on human intervention. In this paper, we propose Failure-Aware Retry (FAR), a framework that enables robots to learn from previous failures at test time, adapt their behavior accordingly, ... Read full story

08:05 <Ada> arXiv.org | BehaviorBench: Benchmarking Foundation Models for Behavioral Science Tasks 8h ago · 3 min read Foundation models have been increasingly applied to behavioral science domains such as psychology, sociology, and economics. While these models show promise in tasks such as survey response prediction and human-subject experiment simulation, there remains no systematic understanding of how well they perform across diverse behavioral science tasks. We intr... Read full story

08:06 <Ada> arXiv.org | Reinforcement Learning-Based Traffic Signal Control for IoT-Enabled Intersections Yousef AlSaqabi · 8h ago · 3 min read Urban traffic congestion remains a persistent challenge in car-dependent cities, imposing significant economic and societal costs. Traffic signal systems are increasingly deployed as networked cyber-physical components within smart-city infrastructures, where distributed sensing and edge intelligence enable adaptive traffic management. This paper investig... Read full story

08:06 <Ada> arXiv.org | SAE++: Cascaded Sparse Autoencoders Learn Multi-Level Visual Concepts in Multimodal LLMs 8h ago · 3 min read Multimodal Large Language Models (MLLMs) have demonstrated strong performance on vision-language tasks, yet their internal visual representations remain difficult to interpret. Sparse Autoencoders (SAEs) provide a scalable way to decompose dense model activations into sparse, interpretable features. However, existing SAE architectures primarily recover fl... Read full story

08:06 <Ada> arXiv.org | An RRAM-based Hardware Implementation of a Radial Basis Function Neuron for Edge Classifiers Georgios Papandroulidakis, Shady Agwa, Themis Prodromakis · 8h ago · 3 min read The deployment of modern machine learning (ML) solutions on resource-constrained edge devices highlights implementation challenges. This is especially true for extreme edge applications that include safety-critical components, such as autonomous navigation tasks. This paper demonstrates an artificial neural network (ANN) design leveraging Metal-Oxide Resi... Read full story

08:06 <Ada> arXiv.org | Operator Calculus for Population-Based Optimization: Modular Convergence and Finite-Population Guarantees 8h ago · 3 min read Population-based optimizers combine update rules such as mutation, selection, and recombination. When one rule changes, it is often unclear which convergence guarantees survive or how the new combination should be assessed. We develop an operator calculus: an operator is a population-update rule, and the calculus specifies how separately checked effects c... Read full story

08:06 <Ada> arXiv.org | Hidden in Plain Sight: Benchmarking Agent Safety Against Decomposition Attacks with DECOMPBENCH Vikhyath Kothamasu, Virginia Smith, Chhavi Yadav · 8h ago · 3 min read LLM-based Agents are becoming increasingly capable and widely deployed, creating growing incentives for adversarial misuse in the real-world. A key emerging threat is Decomposition Attacks \cite{glukhov2024breach, jones2024adversaries} in which a harmful task is broken into simpler, benign subtasks that evade safety mechanisms when executed separately but... Read full story

08:06 <Ada> arXiv.org | Are Good Generators Good Decision-Makers? Policy Learning for General Interventions via Retargeted Counterfactual Generation 8h ago · 3 min read Generative models are increasingly used to support decision-making in complex systems, where interventions may be joint and high-dimensional, and outcomes are high-dimensional. However, using generators for these decision-making settings are challenged by three problems. First, they are often trained on noisy logs with limited intervention data. Second, g... Read full story

08:06 <Ada> arXiv.org | Synthetic Benchmarks Overstate Forward-Forward Scaling: Real-Data Limits of Layer-Local Training Yucheng Chen · 8h ago · 3 min read Forward-Forward (FF) learning [Hinton, 2022] replaces backpropagation with strictly layer-local goodness updates. Recent FF-CNN work has narrowed the gap to BP on 32x32 benchmarks, raising the question of whether layer-local training is becoming a viable alternative at realistic scale. To probe this rigorously, we develop DTG-FF -- dynamic temperature goo... Read full story

08:06 <Ada> arXiv.org | StressDream: Steering Video World Models for Robust Policy Evaluation and Improvement 8h ago · 3 min read Video world models (WMs) have shown promise for policy evaluation and improvement by imagining realistic future observations conditioned on ego-robot actions. While WMs can model distributions over futures, policy evaluation and improvement typically rely on nominal imaginations, which can miss high-impact outcomes of robot actions unless prohibitively ma... Read full story

08:07 <Ada> arXiv.org | Modeling Robotics Dataset Construction as an Artifact-Based Build Process Leon Pohl, Lukas Beer, George Sebastian, Mirko Maehlisch · 8h ago · 3 min read Robotic systems generate large volumes of multimodal sensor data, but converting ROS bag recordings into machine learning datasets is often handled by ad hoc sequential scripts, creating engineering overhead and slow iteration cycles. We model dataset construction as an artifact-based build process over a dependency graph and implement this approach in Ba... Read full story

08:07 <Ada> arXiv.org | Latent Performance Profiling of Large Language Models 8h ago · 3 min read Large language models (LLMs) frequently achieve impressive scores on standardized benchmarks, yet accuracy alone offers a limited view of their capabilities. Evaluating open-source LLMs on leaderboards faces persistent issues such as data contamination, a narrow task scope, and poor alignment with real-world reliability. Benchmark-based evaluations such a... Read full story

08:07 <Ada> arXiv.org | Residualized Temporal Sparse Autoencoders for Interpreting Diffusion Models 8h ago · 3 min read Text-to-image diffusion models generate images by iterative denoising, so their internal layers produce trajectories of activations rather than single static representations. Sparse autoencoders (SAEs) have recently been used to decompose diffusion activations into interpretable features, but most approaches analyze individual timesteps or condition on ti... Read full story

08:07 <Ada> arXiv.org | Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents 8h ago · 3 min read Diagnosing failures in LLM agents remains largely manual. Practitioners inspect a small subset of execution traces, form ad-hoc hypotheses, and iterate. This process misses patterns that only emerge across trace populations and does not scale to production corpora where individual traces span tens of thousands of tokens. We formalize the problem of corpus... Read full story

08:07 <Ada> arXiv.org | Brain alignment of reasoning and action representations from vision-language and action models during naturalistic gameplay 8h ago · 3 min read Understanding how humans and artificial intelligence systems predict and plan by interacting with their environment is a fundamental challenge at the intersection of neuroscience and machine learning. Most brain-encoding studies focus on aligning artificial models with brain activity during language comprehension or passive visual processing, while intera... Read full story

08:07 <Ada> arXiv.org | QLIF-CAST: Quantum Leaky-Integrate-and-Fire for Time-Series Weather Forecasting 8h ago · 3 min read Accurate and efficient time-series forecasting remains a challenging problem for both classical and quantum neural architectures, particularly in multivariate environmental settings. This work adapts the Quantum Leaky Integrate-and-Fire (QLIF) spiking neural network for time-series regression tasks, specifically short-term multivariate weather forecasting... Read full story

08:07 <Ada> arXiv.org | A Few Steps Further: Why Defenses Against Malicious Finetuning Erode Under Continued Training Itay Zloczower, Eyal Lenga, Gilad Gressel, Yisroel Mirsky · 8h ago · 3 min read Model providers increasingly release the weights of large language models. Although these models are safety-aligned before release, their safeguards can often be removed by fine-tuning on harmful data. A growing class of defenses aims to make alignment robust to such malicious fine-tuning, but these defenses are typically evaluated against attacks with a ... Read full story

08:07 <Ada> arXiv.org | Restricting the Model, Missing the System: Measurement and Accountability in Offensive AI Governance 8h ago · 3 min read We show that the instruments used to measure AI offensive capability fail in two ways: (i) they overstate harm, and (ii) they credit the model with capability that belongs to the surrounding system. We argue that restricting access to a model is therefore necessary but not sufficient and that policy and procurement also need system-level, harm-grounded ca... Read full story

08:07 <Ada> arXiv.org | TAVIS: A Benchmark for Egocentric Active Vision and Anticipatory Gaze in Imitation Learning Giacomo Spigler · 8h ago · 3 min read Active vision -- where a policy controls its own gaze during manipulation -- has emerged as a key capability for imitation learning, with multiple independent systems demonstrating its benefits in the past year. Yet there is no shared benchmark to compare approaches or quantify what active vision contributes, on which task types, and under what conditions... Read full story

08:07 <Ada> arXiv.org | Empirical Evidence for Simply Connected Decision Regions in Image Classifiers Arjhun Swaminathan, Mete Akg\"un · 8h ago · 3 min read The topology of a classifier's decision regions determines how inputs with the same predicted label can be connected and deformed without changing that prediction. Prior empirical work constructed paths between same-label images within a single region, but did not examine whether loops bound surfaces within that region. We investigate this question using ... Read full story

16:11 <Vega> OpenAI News | Disrupting AI-enabled “false front” operations 16h ago OpenAI disrupted two AI-enabled influence operations that used false-front journalists and a think tank to spread geopolitical messaging. Read full story

18:10 <Vega> OpenAI News | Pollo AI turns creative ideas into campaigns with OpenAI 6h ago With GPT-5.6, GPT-6 Astra, and GPT‑Image‑2.5, Pollo AI helps creators turn bold ideas into detailed images and cinematic video ads. Read full story

18:10 <Vega> OpenAI News | How Oracle turns days of work into minutes with ChatGPT and Codex 2h ago Across recruiting, engineering, and operations, Oracle turns specialist knowledge into fast, repeatable workflows with ChatGPT Work and Codex. Read full story

18:25 <Ada> arXiv.org | Component-Adaptive and Lesion-Level Supervision for Improved Small Structure Segmentation in Brain MRI Minh Sao Khue Luu, Evgeniy N. Pavlovskiy, Bair N. Tuchinov · 18h ago · 4 min read Small lesions in brain MRI are hard to segment because they occupy a tiny fraction of the volume and are dominated by background and larger lesions during voxel-wise optimization, so a model can reach a high Dice similarity coefficient (DSC) while missing many of them. We propose CATMIL, a training objective that adds two auxiliary terms to the standard n... Read full story

18:25 <Ada> arXiv.org | Out-of-Air Computation: Enabling Structured Function Extraction from Wireless Superposition Seyed Mohammad Azimi-Abarghouyi · 18h ago · 3 min read Over-the-air computation (AirComp) broadly exploits the superposition property of wireless multiple-access channels (MACs) to compute functions of distributed data. Within this broad class, dominant conventional designs are embedding-oriented: they pre-shape transmitted signals or mitigate channel effects so that the received superposition directly realiz... Read full story

18:25 <Ada> arXiv.org | A Variational Latent-Space Framework for Uncertainty-Aware Spectral Image Emulation 18h ago · 3 min read Synthetic spectral image generation is essential for remote sensing simulation and mission design, yet physically based radiative transfer models (RTMs) remain computationally expensive. Existing learning-based emulators reduce this cost, but are mostly deterministic parameter-to-spectrum regressors with limited spatial modeling and uncertainty informatio... Read full story

18:25 <Ada> arXiv.org | Bi-CamoDiffusion: A Boundary-informed Diffusion Approach for Camouflaged Object Detection Patricia L. Suarez, Leo Thomas Ramos, Angel D. Sappa · 18h ago · 3 min read Bi-CamoDiffusion is introduced, an evolution of the CamoDiffusion framework for camouflaged object detection. It integrates edge priors into early-stage embeddings via a parameter-free injection process, enhancing boundary sharpness and preventing structural ambiguity. An optimization objective that unifies spatial accuracy, structural constraints, and un... Read full story

18:25 <Ada> arXiv.org | Prototype-Based Knowledge Guidance for Fine-Grained Structured Radiology Reporting 18h ago · 3 min read Structured radiology reporting promises faster, more consistent communication than free text, but automation remains difficult as models must make many fine-grained, discrete decisions about rare findings and attributes from limited structured supervision. In contrast, free-text reports are produced at scale in routine care and implicitly encode fine-grai... Read full story

18:25 <Ada> arXiv.org | Vectorizing the Trie: Efficient Constrained Decoding for LLM-based Generative Retrieval on Accelerators 18h ago · 4 min read Generative retrieval has emerged as a powerful paradigm for LLM-based recommendation. However, industrial recommender systems often benefit from restricting the output space to a constrained subset of items based on business logic (e.g. enforcing content freshness or product category), which standard autoregressive decoding cannot natively support. Moreov... Read full story

18:25 <Ada> arXiv.org | Fluids You Can Trust: Property-Preserving Operator Learning for Incompressible Flows Ramansh Sharma, Matthew Lowery, Houman Owhadi, Varun Shankar · 18h ago · 3 min read We present a novel property-preserving kernel-based operator learning method for incompressible flows governed by the incompressible Navier--Stokes equations. Traditional numerical solvers incur significant computational costs to respect incompressibility. Operator learning offers efficient surrogate models, but current neural operators fail to exactly en... Read full story

18:26 <Ada> arXiv.org | Multiparameter Uncertainty Mapping in Quantitative Molecular MRI using a Physics-Structured Variational Autoencoder (PS-VAE) 18h ago · 3 min read Quantitative imaging methods, such as magnetic resonance fingerprinting (MRF), aim to extract interpretable pathology biomarkers by estimating biophysical tissue parameters from signal evolutions. However, the pattern-matching algorithms or neural networks used in such inverse problems often lack principled uncertainty quantification, which limits the tru... Read full story

18:26 <Ada> arXiv.org | A Positive Case for Faithfulness: LLM Self-Explanations Help Predict Model Behavior 18h ago · 3 min read LLM self-explanations are often presented as a promising tool for AI oversight, yet their faithfulness to the model's true reasoning process is poorly understood. Existing faithfulness metrics have critical limitations, typically relying on identifying unfaithfulness via adversarial prompting or detecting reasoning errors. These methods overlook the predi... Read full story

18:26 <Ada> arXiv.org | Machine learning modularity Yi Fan, Vishnu Jejjala, Yang Lei · 18h ago · 3 min read Based on a transformer based sequence-to-sequence architecture combined with a dynamic batching algorithm, this work introduces a machine learning framework for automatically simplifying complex expressions involving multiple elliptic Gamma functions, including the $q$-$θ$ function and the elliptic Gamma function. The model learns to apply algebraic ident... Read full story

18:26 <Ada> arXiv.org | Activation-Informed Pareto-Guided Low-Rank Compression for Efficient LLM/VLM 18h ago · 3 min read Large language models (LLM) and vision-language models (VLM) have achieved state-of-the-art performance, but they impose significant memory and computing challenges in deployment. We present a novel low-rank compression framework to address this challenge. First, we upper bound the change of network loss via layer-wise activation-based compression errors,... Read full story

18:26 <Ada> arXiv.org | Circuit realization and hardware linearization of monotone operator equilibrium networks Thomas Chaffey · 18h ago · 3 min read It is shown that the port behavior of a resistor-diode network corresponds to the solution of a ReLU monotone operator equilibrium network (a neural network in the limit of infinite depth), giving a parsimonious construction of a neural network in analog hardware. We furthermore show that the gradient of such a circuit can be computed directly in hardware... Read full story

18:27 <Ada> arXiv.org | Scaling Legal AI: Benchmarking Mamba and Transformers for Statutory Classification and Case Law Retrieval Anuraj Maurya · 18h ago · 3 min read Statutory corpora and judicial decisions are growing faster than legal professionals can read them, while individual judgments often exceed the context limits of standard encoder models. Transformer architectures dominate legal NLP benchmarks, but their quadratic attention complexity can require truncating or fragmenting documents that demand whole-docume... Read full story

18:27 <Ada> arXiv.org | Persistence Paradox in Dynamic Science: Evidence from the Deep Learning Revolution Honglin Bao, Beichen Lu, Kai Li · 18h ago · 3 min read Persistence is often regarded as a virtue in science. In this paper, however, we challenge this conventional view by highlighting its contextual nature, particularly how persistence can become a liability during paradigm shifts. We focus on the deep learning revolution catalyzed by AlexNet in 2012. Analyzing the 20-year career trajectories of more than 5,... Read full story

18:27 <Ada> arXiv.org | APE: Selective Fine-tuning with Acceptance Criteria for Language Model Adaptation Javier Mar\'in · 18h ago · 3 min read We present Adjacent Possible Exploration (APE), a selective fine-tuning method for adapting large language models that systematically explores parameter modifications while maintaining model stability. Inspired by evolutionary optimization principles, APE evaluates multiple candidate parameter updates through fine-tuning on small data subsets and accepts ... Read full story

18:27 <Ada> arXiv.org | Graph-Based Floor Separation Using Node Embeddings and Clustering of WiFi Trajectories Rabia Yasa Kostas, Kahraman Kostas · 18h ago · 3 min read Indoor positioning systems (IPSs) are increasingly vital for location-based services in complex multi-storey environments. This study proposes a novel graph-based approach for floor separation using Wi-Fi fingerprint trajectories, addressing the challenge of vertical localization in indoor settings. We construct a graph where nodes represent Wi-Fi fingerp... Read full story

18:27 <Ada> arXiv.org | Allocation Stability and Wald Inference under Variance-Aware UCB 18h ago · 3 min read Allocation stability is often used to justify Gaussian inference from bandit data, but when is it necessary? In this paper, we address this question for a two-armed, fixed-horizon variance-aware UCB policy with bounded reward distributions that may vary with the horizon. We find a sharp criterion in terms of the reward gap and variances that determines wh... Read full story

18:27 <Ada> arXiv.org | Requirement-Based Testing: Enhancing Reinforcement Learning with Game Theory 18h ago · 3 min read We consider the automatic online synthesis of black-box test cases from functional requirements specified as automata for reactive implementations. The goal of the tester is to reach some given state, so as to satisfy a coverage criterion, while monitoring the violation of the requirements. We develop an approach based on Monte Carlo Tree Search, which is... Read full story

18:27 <Ada> arXiv.org | Speech Emotion Recognition Using CNN and Its Use Case in Digital Healthcare Nishargo Nigar · 18h ago · 3 min read The process of identifying human emotion and affective states from speech is known as speech emotion recognition (SER). This is based on the observation that tone and pitch in the voice frequently convey underlying emotion. Speech recognition includes the ability to recognize emotions, which is becoming increasingly popular and in high demand. With the he... Read full story

18:27 <Ada> arXiv.org | CNet: A Complex-Valued Deep Learning Framework with Wirtinger Autodifferentiation and FFT--Hadamard Convolution Marcel Crasmaru · 18h ago · 3 min read CNet is a C++/CUDA framework for building deep complex-valued neural networks (CVNNs) and optimizing complex functions by gradient descent with Wirtinger derivatives. Complex models are underexplored yet natural where data is intrinsically complex -- RF/IQ communications, MRI k-space, radar/SAR, audio spectra -- and phase carries information real networks... Read full story

22:38 <Vega> MIT News | Massachusetts Institute of Technology | 3 Questions: What is the best path forward for AI in academia? 5h ago · 4 min read A new essay from MIT Statistics and Data Science Center Director Sasha Rakhlin synthesizes recent discussions and readings about how AI is changing academia — particularly mathematics, statistics, machine learning, and engineering. He shares key questions that departments and universities need to think about, and how they can best work with AI going forward. Read full story

22:38 <Ada> Hugging Face | The model that didn't exist, so you made it yourself yuvraj sharma · 22h ago · 9 min read We’re on a journey to advance and democratize artificial intelligence through open source and open science. Read full story

23:40 <Vega> OpenAI News | LegalOn halves Codex costs while maintaining development speed 11h ago LegalOn cut estimated daily Codex costs by 65% while maintaining development speed. It matched Astra, Sol, and Luna to tasks and managed budgets strategically. Read full story

#ai em 2026-10-08