The Archive

Discrete diffusion models are often trained through clean-data prediction, but the prediction can be used in different ways to define the reverse dynamics. In Masked Diffusion Models (MDM) these choices largely coincide, whereas in Uniform Diffusion Models (UDM) they do not. We show that the standard plug-in bridge parameterization for UDM is not optimized by the denoising posterior, but by a leave-one-out posterior that predicts each clean token without using its own noisy observation. This identifies a mismatch between the plug-in ELBO and the usual cross-entropy denoising objective. We cha...

Large language models (LLMs) increasingly excel at mathematical reasoning, but their unreliability limits their utility in mathematics research. A mitigation is using LLMs to generate formal proofs in languages like Lean. We perform the first large-scale evaluation of this method's ability to solve open problems. Our most capable agent autonomously resolved 9 of 353 open Erdős problems at the per-problem cost of a few hundred dollars, proved 44/492 OEIS conjectures, and is being deployed in combinatorics, optimization, graph theory, algebraic geometry, and quantum optics research. A basic age...

While ubiquitous wearable sensors capture a wealth of behavioral and physiological information, effectively transforming these signals into personalized health insights is challenging. Specifically, converting low-level sensor data into representations capable of characterizing higher-level states is difficult due to high phenotypic diversity and variation in individual baseline health, physiology, and lifestyle factors. Moreover, collecting wearable data paired with health outcome annotations is laborious and expensive, and retrospective annotation remains practically unfeasible, contributin...

Random forests are widely used in fields involving sensitive tabular data, but existing approaches to enforcing differential privacy (DP) typically degrade performance to the point of impracticality. In this paper, we introduce Lumberjack, a differentially private random forest algorithm that achieves substantially higher utility by constructing large random decision trees and then applying aggressive, privacy-preserving pruning to retain only sufficiently populated nodes. A key component of our approach is a novel $(\varepsilon,δ)$-DP heavy hitter detection algorithm for hierarchical data, w...

Modern smart grids rely on dense measurement infrastructures, communication links, and intelligent field devices. Although this improves supervision and control, it also increases vulnerability to cyber-physical disruptions. Operators must distinguish physical incidents, such as faults or line disturbances, from malicious actions, such as false data injection or unauthorized command execution. This chapter investigates this problem using the well-known MSU/ORNL Power System Attack Dataset. The proposed method combines machine learning with genetic-algorithm-based feature selection. The object...

Autonomous systems have achieved superhuman performance in isolation or simulation, yet they remain brittle in shared, dynamic real-world spaces. This failure stems from the dominant single-agent paradigm for physical applications, where other actors are ignored or treated as environmental noise, preventing effective coordination. Here we show that multi-agent reinforcement learning provides the essential safety scaffolding required for real-world interaction. Using high-speed quadrotor racing as a high-stakes testbed, we train agents to navigate complex aerodynamic interactions and strategic...

Real-world sensor-based learning systems require uncertainty estimation that is both reliable and computationally efficient. Evidential Deep Learning (EDL) provides single-pass uncertainty estimation by modeling the class probabilities via Dirichlet distributions, where the Dirichlet parameters are predicted by a learned neural network mapping. However, this approach can lead to computational challenges, as Dirichlet expected objectives are more complex than standard supervised learning losses, complicating their analysis and implementation. We address this issue by approximating the objectiv...

Parameter-efficient fine-tuning enables fast personalization of text-to-image diffusion models, but composing multiple custom concepts remains challenging due to representation interference. Existing modular methods either rely on expensive post-hoc fusion or freeze adaptation subspaces, which limit expressiveness and concept fidelity. To address this trade-off, we propose Sequential regularized LoRA (SeqLoRA), a constrained continual learning framework that jointly optimizes both LoRA factors via bilevel optimization. Theoretically, we establish strong convergence guarantees for our algorith...

Decision trees partition the feature space using hard binary thresholds, assigning identical confidence to instances far from a decision boundary and to those directly on it. We introduce ternary decision trees, which augment each split node with an uncertainty zone of half-width delta centered on the optimal threshold. Instances in this zone receive predictions formed by weighted blending of both child subtrees and are flagged as boundary-uncertain, signaling that downstream applications may treat these predictions differently. Crucially, delta is computed locally at each node from statistic...

Shapley and Banzhaf interactions capture the complex dynamics inherent in modern machine learning applications. However, current estimators for these higher-order interactions trade off between speed and accuracy. To overcome this limitation, we introduce ProxySHAP. ProxySHAP reconciles the high sample efficiency of tree-based proxy models with a principled path to consistency via residual correction. On a theoretical level, we derive a polynomial-time generalization of interventional TreeSHAP to compute exact interaction indices for tree ensembles, successfully bypassing exponential tree-dep...

Distillation attacks create a deployment trade-off for model providers: the same outputs that make a model more useful can also make it easier to imitate. We study this trade-off through a minimax game between a utility-constrained teacher and an adaptive student. Our framework yields tractable one-sided response rules: an adaptive evaluation rule in which the student reweights high-value examples, and a teacher-side defense template that suppresses outputs most useful for distillation. From a cheap proxy for example value, we derive Product-of-Experts (PoE), a simple forward-pass-only defens...

Optimization over the intersection of two manifolds arises in a broad range of applications, but is hindered by the coupled geometry of the feasible region. In this paper, we prove that the regularities -- clean intersection and intrinsic transversality -- are equivalent, which yields a tractable projection onto the tangent space of the intersection. Therefore, we propose a geometric method that employs a retraction on only one manifold and updates the iterate along two orthogonal directions. Specifically, the iterates stay on one manifold, and the two directions are responsible for asymptoti...

ChronoMedKG adds temporal reasoning to biomedical knowledge graphs for age-dependent clinical diagnosis; 460K evidence-linked triples.

HarnessAPI unifies LLM tool and HTTP API definitions from single Python source; eliminates duplication across Claude, Cursor agent runtimes.

Compares acoustic emotion recognition vs LLM analysis (Gemini 2.5 Flash) for pathos in political speech; single-speech case study.

State-distribution framework explains post-training (SFT, RL, distillation) beyond loss functions; shows on-policy methods access broader state spaces.

You can buy physical plates to pair with your AI effects. | Photo: Terrence O’Brien / The Verge I'm not sure anyone was really asking for an AI guitar pedal. But it was inevitable that someone would build one. One of the first to take the plunge is Polyend, a well-respected music gear maker with a reputation for building niche, idiosyncratic devices. The company has built grooveboxes around old-school trackers and a multi-effect pedal that you can step sequence. So there was at least some hope that if anyone could do an AI effect pedal right, it would be Polyend. Polyend's Endless is a $299 p...

Reddit discussion on alignment failure mechanisms and early warning signs in AI systems.

Establishes near-optimal approximation and generalization bounds for Multiple Neural Operators in multi-task learning with shared representations.

Covariance matching in Gaussian DDPMs improves path-KL divergence from Ω(1/T) to better scaling; improves classifier guidance stability.

Tests OpenAI, Anthropic, DeepSeek, xAI models for conflict-context failures: false atrocity equivalence, genocide denial, ethnic slur misrecognition.

Sparse autoencoder audit identifies 146 GPT-2 Small features correlating with IOI task failures; strongest predictor is 'cryptographic keys' feature.

Adapts audio diffusion models for interactive live music on consumer hardware via block-wise outpainting; avoids discrete-AR compute requirements.

Parametric modular logic programs extend ASP with subprogram parameters and intensionality; connects clingo collective control to traditional ASP.

AnyMo framework enables human motion recognition across heterogeneous wearable IMU sensors via geometry-aware setup-agnostic modeling.

Study of 75,898 API calls shows LLMs exhibit accumulated message effect bias when evaluating sequential items in single conversations.

Reddit speculation on Gorgon Halo vs Strix Halo memory bandwidth; claims 6.7% speedup insufficient for LLM inference, awaiting Medusa Halo.

Hierarchical offline GCRL framework using relativised options and representational abstraction to reuse experience across state-goal symmetries.

Qualitative study of 24 professionals at large tech firm shows AI adoption blurs role boundaries and transforms informal workplace culture.

ToaST tokenization method optimizes compression via recursive split trees and integer programming, decoupled from fixed vocabularies.