Frontier Research Intelligence
Preparing the latest research view.
Browse
Search and filter the research library by collection and research area. Results are URL-driven for shareable views.
Scaling inference compute, by generating many parallel attempts per problem, is a costly but reliable lever for improving language model capabilities. By default these attempts are generated independently, wasting inference compute on redundant solutions. This waste seems unavoidable. After all, independence is what makes parallel sampling trivial to scale. However, this tradeoff is not fundamental: there is a ric...
Background The field of implantable Brain Computer Interfaces (iBCIs) is rapidly advancing, with individuals with amyotrophic lateral sclerosis (ALS) as key beneficiaries. However, ALS related cortical degeneration may impair iBCI effectiveness. This study investigated whether structural magnetic resonance imaging (MRI) and functional MRI (fMRI) metrics are associated with the quality of electrocorticography (ECoG...
The four channel teleoperation architecture is a well established framework for achieving transparency in bilateral systems. However, its performance in human scale teleoperation is limited by high inertia, modeling challenges, and reliance on noisy and costly force/torque sensors. This paper introduces a sensorless four channel architecture based on inverse dynamics modeling. The controller is implemented and val...
Classification of motor imagery (MI) tasks through EEG is valuable in brain computer interfacing and rehabilitation engineering. EEG channels selection for MI task classification is well discussed problem and is challenging due to its combinatorial nature. Most of the existing methods are subject and task dependent. This paper introduces a subject independent EEG channel selection. The proposed approach consists o...
Video predictive models are emerging as a powerful paradigm in robotics, offering a promising path toward task generalization, long horizon planning, and flexible decision making. However, prevailing approaches often operate on 2D video sequences, inherently lacking the 3D geometric understanding necessary for precise spatial reasoning and physical consistency. We introduce a Structured 4D Latent Predictive Model,...
Memory expertise is a learned skill: knowing what to encode, when to retrieve, and how to organize knowledge a capacity known in cognitive science as metamemory. We bring this perspective to LLMs by treating memory management as a trainable skill. We promote file system operations to first class memory actions alongside task actions, letting the model itself decide how to manage its memory. This memory skill impro...
When should an AI system's answer be trusted? Formal proof assistants offer certainty but cannot reach most of the problem distribution; scalar LLM judges offer coverage but produce opaque scores that cannot be audited after the fact and are subject to the same coherence issues as any LLM. We present Theoria, a verification architecture that closes this gap. A candidate solution is rewritten into a sequence of typ...
This paper studies real time robust optimal control for uncertain nonlinear systems, where linear time varying (LTV) approximations make planning tractable but require sound linearization error bounds (LEBs) to guarantee robust constraint satisfaction. We develop tight, differentiable, GPU parallel LEBs for LTV approximations of nonlinear and neural network (NN) dynamics. For analytic dynamics, we introduce path b...
Digital watermarking secures electroencephalogram (EEG) data in distributed Brain–Computer Interface (BCI) environments. However, existing single domain deep learning schemes struggle to maintain robustness against clinical signal processing attacks due to EEG’s joint time–frequency nature. We introduce the Dual Stream Frequency–Time Attention Network (DS FTAN), utilizing an adaptive Spectral Gating Mechanism to e...
Fear related disorders—including post traumatic stress disorder (PTSD), specific phobias, and generalized anxiety disorder—are among the most prevalent and treatment resistant psychiatric conditions. Despite decades of molecular and cellular advances, systemic pharmacological interventions and approaches targeting isolated brain regions have had limited therapeutic capacity, particularly in refractory patients. Th...
For predictive models, the often reported performance metrics are the loss and accuracy. In synchronous Brain Computer Interface (BCI) systems, these metrics are informative for most BCI paradigms; however, for Event Related Potential (ERP) applications the spelling rate, which measures the number of characters correctly selected is more important as it influences the estimation of information transfer rate (ITR) ...
Vision language action (VLA) models across robot embodiments require high quality observation action supervision to learn deployable action distributions, yet scaling such robot data remains difficult, especially for high DoF humanoids. Teleoperation provides controller aligned supervision, while human egocentric videos capture diverse bimanual manipulation but do not directly provide executable robot actions. We ...
The tendency of large generative models to memorize training data makes sample verification critical for privacy auditing and copyright enforcement. Current membership (MIA) and dataset inference (DI) attacks often rely on one shot generations, which yield weak signals and limited sensitivity across modalities. Inspired by Model Autophagy Disorder (MAD), we introduce MADreMIA, a model agnostic framework that enhan...
Agentic reinforcement learning requires assigning credit to environment facing actions such as searches, clicks, edits, navigation commands, and object interactions. Standard GRPO uses the final verifier outcome as a uniform advantage over all action tokens. This outcome signal is useful but structurally incomplete: it punishes useful exploration in failed rollouts and reinforces redundant or regressive actions in...
Language models are increasingly taught from synthetic question answer (QA) supervision: a model generates questions about a document, answers them from the same text, and the resulting pairs are used to fine tune, distill, or compress knowledge into another model. We show that this generation step is not neutral preprocessing. It is an implicit policy that both selects which evidence becomes training signal and d...
Generalist robot policies learn a diverse repertoire of behaviors from large scale pretraining. In principle, this makes them excellent priors for downstream adaptation via reinforcement learning (RL). In practice, however, standard RL methods leveraging this prior optimize directly over robot actions, requiring the base policy's action distribution to be close to that of a performant policy from the start. This a...
Metacognition is a critical component of intelligence that describes the ability to monitor and regulate one's own cognitive processes. Yet LLMs exhibit systemic deficiencies in key metacognitive faculties: they hallucinate with high confidence, fail to recognize knowledge boundaries, and misrepresent their internal uncertainty undermining trustworthiness and reliability. Since monitoring task performance and adap...
We study agentic code generation in Dafny, where a model must generate both executable code and the proof artifacts for verification. We present AxDafny, a verifier guided repair framework that iteratively generates implementations, invariants, assertions, and termination arguments. We also introduce LiveCodeBench Pro Dafny (LCB Pro Dafny), a benchmark of 250 competition style programming problems translated into ...
Physiotherapy has traditionally been delivered face‑to‑face in clinical settings, but emerging technologies like telerehabilitation, wearable sensors, virtual/augmented reality, robotic exoskeletons, and artificial intelligence (AI) are rapidly changing this model. This narrative review of 42 peer‑reviewed, open‑access articles published between 2018 and 2026 examines their implications for equitable health servic...
Latent world models enable planning from high dimensional observations by predicting future states in a compact latent space. However, these models are typically kept frozen at test time: when their predictions become inaccurate, planning can fail, especially under test time distribution shift. To address this, we propose AdaJEPA, an adaptive latent world model that performs test time adaptation within the closed ...
